Dokumentenidentifikation EP1334375 07.10.2004
EP-Veröffentlichungsnummer 0001334375
Anmelder WesternGeco Ltd., Gatwick, West Sussex, GB
Erfinder BERNTH, Henrik, Cambridge, Cambridgeshire CB3 9BD, GB;
DAUBE, Francois, Houston, US
Vertreter derzeit kein Vertreter bestellt
DE-Aktenzeichen 60105309
Vertragsstaaten AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LI, LU, MC, NL, PT, SE, TR
Sprache des Dokument EN
EP-Anmeldetag 09.11.2001
EP-Aktenzeichen 019807601
WO-Anmeldetag 09.11.2001
PCT-Aktenzeichen PCT/GB01/04977
WO-Veröffentlichungsnummer 0002039144
WO-Veröffentlichungsdatum 16.05.2002
EP-Offenlegungsdatum 13.08.2003
EP date of grant 01.09.2004
Veröffentlichungstag im Patentblatt 07.10.2004
IPC-Hauptklasse G01V 1/30
IPC-Nebenklasse G01V 1/36   


The present invention relates to a method of processing seismic data and provides a technique for computing a stacked line by interpolation between known stacks.

Seismic data are collected using an array of seismic sources and seismic receivers. The data may be collected on land using, for example, explosive charges as sources and geophones as receivers. Alternatively, the data may be collected at sea using, for example, airguns as sources and hydrophones as receivers. After the raw seismic data have been acquired, the reflected signals (known as traces) received by each of the receivers as a result of the process of actuation of a seismic energy source are processed to form a subsurface image. The processing includes the steps of accounting for the separation (known as offset) between sources and receivers and summing related traces together to increase signal/noise ratio (a process known as stacking).

Figure 1 of the accompanying drawings schematically illustrates an idealised source and receiver arrangement arranged along a line. First, second and third sources 1, 2 and 3, respectively, cooperate with first, second and third receivers 4, 5 and 6, respectively. The sources and receiver are arranged about a common midpoint (CMP) 7 for the source/receiver pairs 1, 6; 2, 5; 3, 4. Seismic energy produced from the actuation of each of the sources 1, 2 and 3 is reflected from partial reflectors such as 9 and received by each of the receivers 4, 5 and 6. The travel time of the energy from a source to a receiver increases with increasing distance (offset) between the source and the receiver. The travel time is also a function of the depth of the reflectors and of the velocity of propagation of the signal within the subsurface formations.

Figure 2 of the accompanying drawings illustrates the travel time for the situation shown in Figure 1, as the offset increases. The round trip travel time with respect to offset for each of the reflectors defines a curve. In this simplified situation, the curve can be accurately defined by: t2(offset)=(offset)2/(velocity)2+t2(zerooffset) where t is the round trip travel time, offset is the distance between source and receiver and velocity is the speed of propagation of seismic signals within the subsurface formations.

During the processing of the seismic survey data, the traces are assigned to their respective common midpoints such that the geology beneath the line of sources and receivers can be probed at a plurality of positions. A velocity analysis is then performed for each common midpoint and indeed for each reflector 9. This is achieved by specifying a range of hyperbolas, as defmed in the above equation, related to a range of velocities and computing the reflection amplitude along all specified hyperbolas. The seismic traces for a plurality of offsets are then converted in accordance with the hyperbolas to equivalent traces having zero offset and the traces are then summed (stacked). The resulting amplitudes at zero offset are examined to determine which hyperbola gives the best result for each of the reflectors of each common midpoint. Figure 3 of the accompanying drawings shows a typical example of velocity analysis at point i, where the velocity function selected by the user varies between a range of known velocities functions.

Furthermore, as described in the article 'MIGRATION-VELOCITY ANALYSIS FOR TI AND ORTHORHOMBIC BACKGROUND MEDIA' L.T. IKELLE, revue de l'institut français du petrol, editions techniq. Paris, FR, vol. 53, No. 5, sept 1998, p9. 571-584, there is provided a seismic data processing nethod in which a number of seismic stacks are precomputed for known velocity fields which are chosen to span the range of velocities of interest, and the stacks are then arranged in a 3D volume, using time and position as first dimensions and the index of the velocity field as the last dimension, to provide a seismic line stacked for a velocity field of interest.

Once a velocity function has been analyzed for a common midpoint, the seismic data related to the common midpoint are then corrected to zero offset according to the previous equation and then stacked for that particular common midpoint. The stacked trace has an improved signal-noise ratio compared to the traces recorded at the receivers. That process, repeated at each of the common midpoints of the line, produces a stacked seismic line that gives an indication of the geology of the line. The quality of the stacked line is directly related to the quality of the velocity field used for stacking. Stacking a line is a CPU intensive process that necessitates the use of large and powerful machines, especially if it is to be done in real time.

The present invention replaces the method of conventional stacking with a technique based on interpolation that can be performed very quickly on modem graphics computers, as set forth in the appended claims.

The invention will be further described, by way of example, with reference to the accompanying drawings, in which:

  • Figure 1 illustrates diagrammatically a seismic source/receiver arrangement of known type;
  • Figure 2 is a graph illustrating travel time against offset;
  • Figure 3 illustrates various velocity functions as velocity against time;
  • Figure 4 is a block schematic diagram illustrating a computer for performing a method constituting an embodiment of the invention;
  • Figure 5 is a flow diagram illustrating a method of processing seismic data constituting an embodiment of the invention;
  • Figure 6 illustrates a three dimensional array of data stored in the memory of the computer shown in Figure 4; and
  • Figures 7 and 8 illustrate an example of interpolation performed by the method shown in Figure 5.

The computer shown in Figure 4 comprises a central processing unit (CPU) 10 provided with an input arrangement 11 and an output arrangement 12, for example including a display for displaying the results of the processing performed by the CPU 10. The computer has a program memory 13 which contains a computer program for controlling the operation of the CPU 10 to perform a seismic data processing method as described hereinafter. The computer also has a scratchpad memory 14 for temporarily storing data during operation of the CPU 10, including a three dimensional (3D) memory which cooperates with graphics processing software in the program memory 13 to perform graphics processing including interpolation. The computer therefore functions as a graphics computer, such as a Sun or Silicon Graphics workstation or a high end PC having sufficient memory to store a large "cube" of data and a 3D graphics card with texture mapping which supports the OpenGL language from Silicon Graphics to perform the interpolation.

The method performed by the computer shown in Figure 4 is illustrated in Figure 5 and begins with a step 20 which precomputes a plurality of seismic stacks from seismic data supplied to the computer for a plurality of known velocity fields or functions V1,....,Vn. In particular, the computer precalculates a stack for each common midpoint (CMP) and for each velocity field Vi. The known velocity functions are selected so as to span a range of velocities of interest. These known velocity functions may be selected in any suitable way and may be selected essentially arbitrarily or may be based on knowledge or experience associated with the seismic data being processed. For example, one of the velocity functions may be selected in accordance with any suitable criteria and the remaining known velocity functions may be equal to the product of the known velocity function and a set of coefficients. For example, the velocity functions may differ from each other by fixed percentages or fixed ratios to provide evenly spaced velocity functions spanning the range of velocities of interest.

The stacks formed in the step 20 are then arranged as a 3D array of stacks in the memory 14 in a step 21. For example, as shown in Figure 6, the stacks are arranged in a rectangular 3D array as a cube of data with the vertical downward dimension representing increasing time, the right hand horizontal axis representing common midpoint number, and the depth axis into the plane of Figure 6 representing the velocity function index with the velocity functions being indexed in increasing order of velocity.

In the 3D space containing the cube of stacked seismic data, any selected or chosen velocity function or field is represented by a velocity surface S as illustrated in Figure 4 (provided that the values of the selected surface lie within the range of velocities spanned by the known velocity functions, for example between the extreme or end functions V1 and Vn). As shown by the step 22 in Figure 5, a velocity function is selected for further processing of the seismic data by a user and this determines the velocity surface S onto which the seismic data can be projected.

A step 23, for example performed within the graphics card, performs interpolation effectively so as to define a "stacked" line of seismic data based on interpolating onto the surface S from the individual samples of the stacks within the data cube which surround each point of the surface S. Although this processing does not yield a true stack, it provides a representation thereof and can be performed relatively quickly, for example in real time, using relatively modest hardware and software. A specific example of an interpolation technique will be described hereinafter.

A step 24 outputs the seismic line, for instance by displaying it on a display of the output arrangement 12 of the computer. A step 25 determines whether a new velocity function has been selected. Until a new function is selected, the output for the existing velocity function S remains available. A user may therefore examine the result of the processing in the form of the stacked line and may decide select a new function, for example by changing the velocity function or chosing a different function. When a new velocity function is selected, control returns to the step 23, which performs a fresh interpolation based on the new function.

Although the stacked traces shown in Figure 6 are illustrated as continuous traces, they are in fact sampled and digitised so that each of the stacks comprises a plurality of digital codes representing the instantaneous sampled amplitude at discrete time points. The cube or space of 3D data is thus effectively divided into a plurality of cells, each of which is cuboidal and has at its vertices eight stack samples S(i,j,k),...., S(i+1, j+1,k+1) as illustrated in Figure 7. The coordinate axis are shown again in Figure 7 with the common midpoint (CMP) index increasing towards the right in the horizontal or x dimension, time increasing downwardly in the vertical or y dimension, and velocity function index increasing in the depth or z dimension into the plane of Figure 7. The sample which occupies the top front left vertex of the cell is labelled as S(i,j,k) and the remaining seven samples at the other verticies are labelled in accordance with the convention of the axes as described hereinbefore. A sample Sp is illustrated within the cell at a point on the velocity surface S at which it is desired to calculate the "output sample" for the selected velocity function. Thus, the cell illustrated in Figure 7 is one of the cells intersected by the velocity surface S and the value of each sample Sp within a respective cell is calculated by the graphics card by interpolation.

Figure 8 illustrates the position 30 at which the sample Sp is to be calculated within the same cell as illustrated in Figure 7. The point 30 may be anywhere within the cell, including the surfaces, edges and verticies thereof as well as internally within the volume of the cell. Without any loss of generality, the position of the point 30 can be represented by the distances a, ....,f from the various faces of the cuboidal cell. Thus, the point 30 is at a distance a from the front face and b from the rear face, a distance c from the top face and d from the bottom face, and a distance e from the left face and f from the right face, where any of the distances may be zero so that the position of the point 30 can be specified anywhere within the cell including on the external surface thereof.

The graphics card within the computer shown in Figure 4 performs a linear interpolation in order to calculate the value or amplitude of the sample Sp from the samples S(i,j,k),....,S(i+1, j+1, k+1) in accordance with a multilinear (in this case trilinear) interpolation which may be represented as follows: Sp=(a.d.s.S(i,j,k+1)+b.d.f.S(i,j,k)+a.c.f.S(i,j+1,k+1) +b.c.f.S(i,j+1,k)+a.d.e.S(i+1,j,k+1)+b.d.e.S(i+1,j,k)+a.c.e.S(i+1,j+1,k+1)+b.c.e.S(i+1,j+1,k))/((a+b)(c+d)(e+f))

The interpolation is performed for every cell of the cube of data intersected by the velocity surface S and thus provides a representation or approximation of a stacked line. This may be repeated for a plurality of lines to give a 3D representation of the subsurface structure of the earth represented by the seismic data.

Any suitable interpolation method may be performed within the step 23. For example, any suitable software, such as existing graphics card software, maybe used.

It is thus possible to provide a technique which allows a good representation of a stacked line to be derived relatively quickly and with relatively inexpensive hardware and software. This may be used, for example, in real time. Also, different velocity functions can be tried relatively quickly in order to allow a user to choose the best such function to fit the seismic data. When an optimum velocity function has been selected, it may be used for re-stacking of the seismic traces.

  1. Verfahren zur Verarbeitung von seismischen Daten, wobei das Verfahren die folgenden Schritte umfaßt:
    • (a) Vorausberechnung einer Vielzahl von seismischen Stacks in einer Vielzahl von Positionen und für eine Vielzahl von vorbestimmten Geschwindigkeitsfunktionen, die einen Bereich von interessierenden Geschwindigkeiten überspannen, aus den seismischen Daten;
    • (b) Anordnen der Stacks als dreidimensionale Anordnung, wobei Zeit, Position und Geschwindigkeitsindex als die drei Dimensionen der Anordnung fungieren; und
    • (c) Definieren einer 3D-Geschwindigkeitsfläche innerhalb des Bereichs der interessierenden Geschwindigkeiten, wobei die Geschwindigkeitsfläche den Bereich der vorbestimmten Geschwindigkeitsfunktionen schneidet;

         wobei das Verfahren dadurch gekennzeichnet ist, daß der Schritt (b) den folgenden Schritt umfaßt: Anordnen der Stacks als dreidimensionale Anordnung in einem Speicher eines Grafikcomputers; und daß es den folgenden weiteren Schritt umfaßt:
    • (d) Verwenden eines Grafikprogramms des Computers, um aus der Anordnung der Stacks eine seismische Linie abzuleiten, die seismische Daten darstellt, die in Form eines Stacks für die definierte Geschwindigkeitsfläche dargestellt werden, indem anhand einer Menge von Werten in den Stacks, die jeden Punkt der gewählten Geschwindigkeitsfläche umgeben, interpoliert wird.
  2. Verfahren nach Anspruch 1, wobei die Positionen gemeinsame Mittelpunkte der seismischen Daten umfassen.
  3. Verfahren nach Anspruch 1 oder 2, wobei mindestens einige der vorbestimmten Geschwindigkeitsfunktionen beliebig gewählt werden.
  4. Verfahren nach Anspruch 3, wobei die vorbestimmten Geschwindigkeitsfunktionen eine erste Funktion und eine Vielzahl von zweiten Funktionen umfassen, von denen jede gleich dem Produkt aus der ersten Funktion und einem jeweiligen Koeffizienten ist.
  5. Verfahren nach Anspruch 4, wobei die Koeffizienten im wesentlichen gleich beabstandet sind.
  6. Verfahren nach einem der Ansprüche 1 bis 5, wobei die Anordnung eine rechteckige Anordnung ist.
  7. Verfahren nach einem der vorhergehenden Ansprüche, wobei die Interpolation eine lineare Interpolation ist.
  8. Verfahren nach Anspruch 7, wobei die Interpolation eine multilineare Interpolation ist.
  9. Verfahren nach Anspruch 8, wobei die Interpolation eine trilineare Interpolation ist.
  10. Computer, der programmiert ist, um ein Verfahren nach einem der vorhergehenden Ansprüche durchzuführen.
  11. Programm zur Ausführung auf einem Computer, um ein Verfahren nach einem der Ansprüche 1 bis 9 durchzuführen.
  12. Speichermedium, das ein Programm nach Anspruch 11 enthält.
  1. A method of processing seismic data, the method comprising the steps of:
    • (a) precomputing from the seismic data a plurality of seismic stacks at a plurality of positions and for a plurality of predetermined velocity functions which span a range of velocities of interest;
    • (b) arranging the stacks as a three dimensional array with time, position and index of velocity function as the three dimensions of the array; and
    • (c) defining a 3D velocity surface within the range of velocities of interest, the velocity surface intersecting the range of the predetermined velocity functions;

         the method being characterised in that step (b) comprises arranging the stacks as the three dimensional array in a memory of a graphics computer; and in that it comprises the further step of:
    • (d) using a graphics program of the computer to derive from the array of stacks a seismic line representing seismic data stacked for the defined velocity surface by interpolating from a set of values in the stacks surrounding each point of the selected velocity surface.
  2. A method as claimed in claim 1, in which the positions comprise common midpoints of the seismic data.
  3. A method as claimed in claim 1 or 2, in which at least some of the predetermined velocity functions are selected arbitrarily.
  4. A method as claimed in claim 3, in which the predetermined velocity functions comprise a first function and a plurality of second functions, each of which is equal to the product of the first function and a respective coefficient.
  5. A method as claimed in claim 4, in which the coefficients are substantially evenly spaced.
  6. A method as claimed in any one of claims 1 to 5, in which the array is a rectangular array.
  7. A method as claimed in any preceding claim, in which the interpolation is a linear interpolation.
  8. A method as claimed in claim 7, in which the interpolation is a multilinear interpolation.
  9. A method as claimed in claim 8, in which the interpolation is a trilinear interpolation.
  10. A computer programmed to perform a method as defined in any one of the preceding claims.
  11. A program for running on a computer to perform a method as defined in any one of claims 1 to 9.
  12. A storage medium containing a program as claimed in claim 11.
  1. Procédé de traitement de données sismiques, le procédé comprenant les étapes de:
    • (a) précalcul, à partir des données sismiques, d'une pluralité d'empilements sismiques en une pluralité de positions et pour une pluralité de fonctions de vitesse prédéterminées qui englobent une plage de vitesses digne d'intérêt;
    • (b) agencement des empilements en tant que réseau tridimensionnel en fonction du temps, de la position et de l'index de fonction de vitesse en tant que les trois dimensions du réseau; et
    • (c) définition d'une surface de vitesse 3D à l'intérieur de la plage de vitesses digne d'intérêt, la surface de vitesse intersectant la plage des fonctions de vitesse prédéterminées,

         le procédé étant caractérisé en ce que l'étape (b) comprend l'agencement des empilements en tant que réseau tridimensionnel dans une mémoire d'un ordinateur de graphique et en ce qu'il comprend en outre l'étape supplémentaire de:
    • (d) utilisation d'un programme de graphique de l'ordinateur pour dériver, à partir du réseau d'empilements, une ligne sismique qui représente des données sismiques empilées pour la surface de vitesse définie en réalisant une interpolation à partir d'un jeu de valeurs dans les empilements entourant chaque point de la surface de vitesse sélectionnée.
  2. Procédé selon la revendication 1, dans lequel les positions comprennent des points médians ou intermédiaires communs des données sismiques.
  3. Procédé selon la revendication 1 ou 2, dans lequel au moins certaines des fonctions de vitesse prédéterminées sont sélectionnées de façon arbitraire.
  4. Procédé selon la revendication 3, dans lequel les fonctions de vitesse prédéterminées comprennent une première fonction et une pluralité de secondes fonctions dont chacune est égale au produit de la première fonction et d'un coefficient respectif.
  5. Procédé selon la revendication 4, dans lequel les coefficients sont espacés de façon sensiblement uniforme ou régulière.
  6. Procédé selon l'une quelconque des revendications 1 à 5, dans lequel le réseau est un réseau rectangulaire.
  7. Procédé selon l'une quelconque des revendications précédentes, dans lequel l'interpolation est une interpolation linéaire.
  8. Procédé selon la revendication 7, dans lequel l'interpolation est une interpolation multilinéaire.
  9. Procédé selon la revendication 8, dans lequel l'interpolation est une interpolation trilinéaire.
  10. Ordinateur programmé pour réaliser un procédé selon l'une quelconque des revendications précédentes.
  11. Programme pour un déroulement sur un ordinateur pour réaliser un procédé tel que défini selon l'une quelconque des revendications 1 à 9.
  12. Support de stockage qui contient un programme selon la revendication 11.

A Täglicher Lebensbedarf
B Arbeitsverfahren; Transportieren
C Chemie; Hüttenwesen
D Textilien; Papier
E Bauwesen; Erdbohren; Bergbau
F Maschinenbau; Beleuchtung; Heizung; Waffen; Sprengen
G Physik
H Elektrotechnik



Patent Zeichnungen (PDF)

Copyright © 2008 Patent-De Alle Rechte vorbehalten. eMail: