A method of seismic wave impedance diffusive inversion

Abstract

The present application relates to a kind of seismic wave impedance diffusion inversion method and device, its essence is a kind of from well site to surrounding diffusion layered structure and attribute inversion method.Core is to use well logging curve to carry out layering etc., then carry out well position inversion, simultaneously, to make the result obtained by well logging interpretation and actual seismic record difference minimum, need to be appropriately disturbed by the square wave type calibration layering curve obtained from well logging interpretation, generate hundreds of reflection coefficient curves, to generate synthetic seismic record, find the synthetic record with actual seismic record difference minimum, the horizon etc. corresponding to this best synthetic record is regarded as the best seed solution and diffuses to all around, during diffusion process, by adding time window or horizon to constrain vertical propagation interval, to obtain the inversion optimum value of entire data set, specific flow as shown in Figure 1.The present application improves the precision of thin layer boundary delineation, and has higher potential for thin layer resolution.

Description

Technical Field

[0001] This invention is applied to geophysics, seismic exploration and development, and especially to the prediction of thin-layer property values. Background Technology

[0002] In the field of seismic exploration, traditional 3D seismic inversion has been widely used for the detailed description of various oil and gas reservoirs. However, the inherent resolution bandwidth of seismic data limits the resolution of reservoir descriptions, especially for identifying the boundaries and interiors of thin sand-mud interbedded reservoirs. Seismic inversion requires prior geological information and well constraints, after which seismic information is incorporated into the inversion results. Traditional deterministic inversion can obtain a globally optimal solution for rock properties and their spatial variations, but this optimal solution is also a least-squares solution, and inherent resolution problems remain for thin-layer identification. Stochastic inversion provides a set of simulated implementation schemes, each of which is a possible scheme with a high-resolution description of the target reservoir. However, in stochastic inversion, the high resolution of each inversion is purely the result of geostatistical simulation.

[0003] To overcome these challenges, we propose a high-resolution inversion solution that goes beyond simple simulation—a method and apparatus for seismic impedance diffusion inversion. This has guiding significance for the exploration and development of thin interbedded clastic sandstone and mudstone reservoirs. Summary of the Invention

[0004] The purpose of this invention is to address the limitations of deterministic and stochastic inversion methods. While a large number of wells are required to obtain satisfactory inversion results, the reliability of the results is generally low when the number of wells is small, and inherent resolution issues exist. Therefore, this invention proposes a method and apparatus for seismic impedance diffusion inversion. This diffusion inversion method diffuses layered structures and properties from the well location outwards. This method minimizes the difference between the seismic solution obtained from rock property solutions and the synthesized solution. If the seismic inversion solution from the available well location is high-resolution, the diffusion inversion method can diffuse this high-resolution solution throughout the entire seismic data. To implement this method, thin-layer interpretation of the logging curves is first required to determine the layer thickness, impedance properties, and Poisson's ratio, etc. Then, thickness and property inversion is performed at the well location. In this step, since the results obtained from the well logging interpretation may differ from the seismic records around the well, to reduce this difference, the rock properties and layer thickness interpreted by the well logging need to be appropriately perturbed, generating hundreds of perturbed reflection coefficient curves. Then, a synthetic record is generated based on these curves, and the synthetic record with the smallest difference from the seismic record is selected for the next step, thus minimizing the difference between the seismic record and the synthetic record. Finally, the optimal solution at the well location is used as a seed to diffuse to other traces of the entire seismic data. A time window can be added during diffusion to constrain the vertical diffusion distance. This method has advantages such as being fast, reliable, and not requiring a low-frequency model.

[0005] To achieve the above objectives, the present invention provides a method and apparatus for seismic wave impedance diffusion inversion, the method comprising:

[0006] Well selection and stratification are performed in wells that align with prior knowledge.

[0007] Well location inversion is performed, and by perturbing parameters such as layer thickness and rock properties obtained from well logging interpretation, a synthetic record with the smallest difference from the seismic record around the well is obtained, and parameters such as location, thickness and properties are optimized.

[0008] The optimal solution at the well location is used as a seed solution to propagate and invert to the surrounding area;

[0009] The diffusion process involves selecting seismic traces for inversion from the well outwards. The inversion result is the solution with appropriate adjustments to the seed points. This process must satisfy the condition that the adjustment range is not too large and that the difference between the synthetic record and the seismic record is small. This process is called the diffusion inversion process.

[0010] When diffusion inversion values ​​originating from one well meet diffusion inversion values ​​originating from another well at a certain seismic trace, the diffusion inversion values ​​from both sides are averaged at the meeting point. Attached Figure Description

[0011] Figure 1This is a flowchart of a method and apparatus for seismic wave impedance diffusion inversion.

[0012] Figure 2 A schematic diagram illustrating the specific implementation process of a method and apparatus for seismic wave impedance diffusion inversion.

[0013] Figure 3 This is a comparison diagram of the actual wave impedance profile with the traditional inverted wave impedance profile and the diffuse inverted wave impedance profile.

[0014] Figure 4 This is a comparison diagram of the true Poisson's ratio profile with the traditional inverted Poisson's ratio profile and the diffusion inverted Poisson's ratio profile. Detailed Implementation

[0015] The specific process and implementation method of the present invention will be described in more detail below. The concept of the present invention, the effects produced, and the predictive potential for thin reservoirs will be fully described below with reference to specific examples.

[0016] According to the present invention, a method and apparatus for seismic wave impedance diffusion inversion are provided, the method comprising the following detailed steps:

[0017] Step 1: Select certain wells as seed wells for diffusion inversion. At these well locations, interpret and calibrate the wave impedance and Poisson's ratio attribute curves of thin interbedded layers. The original well locations have high-resolution logging curves. Based on these logging curves, interpret the sandstone and mudstone layers. After layering, calculate the average value of each layer's curve, thus forming the initial square wave calibration layering curves (see...). Figure 1 );

[0018] Step 2: Well Location Inversion; The synthetic trace generated using the calibrated square wave curve is compared with the seismic trace. The interface position and mean value of the square wave curve are adjusted to make the synthetic trace closer to the seismic trace. The adjusted square wave curve is the seed solution for this well location. The process of obtaining the seed solution is as follows: Using a global search algorithm, based on the original square wave layering, the layer thickness, wave impedance attribute curves, and Poisson's ratio attribute curves are appropriately perturbed to obtain hundreds of reflection coefficient curves. Then, a synthetic record is generated based on these results to find the optimal synthetic solution with the smallest difference from the specific seismic record around the selected well location. The final optimal synthetic result corresponds to the new optimal square wave curve layering, the adjusted value, and the well location adjustment, which is the seed solution for the well location. The expression for finding the minimum difference between the synthetic record and the seismic record is as follows:

[0019]

[0020] In the formula: To find the sign of the minimum value of the expression;

[0021] It is the symbol for the second norm in mathematics;

[0022] Seed solution for reflection coefficient;

[0023] For the extracted seismic wavelet;

[0024] The reflection coefficient before the disturbance;

[0025] For earthquake records;

[0026] The damping factor is the weight of the two L2 norms in the minimum difference calculation expression;

[0027] Step 3: Diffusion Inversion from Well Outwards; First, using the seed solution at the well location as a reference, select any adjacent well trace. The seed solution at this point is considered as R′ in the expression above. Similarly, use the expression above to obtain a new seed solution R for any adjacent well trace location. This expands the high-resolution original seed solution to trace locations around the well. This process continues throughout the diffusion process. For each new trace location, the solution references the seed solutions at adjacent locations, maintaining the same number of layers, thus diffusing to the entire seismic data volume (see...). Figure 2 In addition, by adding a time window or stratigraphic level to the vertical direction of the seismic trace to constrain the range of propagation interval, the thickness range of the reservoir can be determined.

[0028] Within the study area, using the aforementioned methodology, seismic profiles were generated by convolution of the extracted wavelet and the real model. Rock properties were extracted at five different locations, and five virtual wells were established. Traditional inversion used interpolation from these five wells to establish a low-frequency model, while diffusion inversion used these five wells as seed points to diffuse the data set. The wave impedance / Poisson's ratio obtained from both traditional and diffusion inversions was compared with the actual wave impedance / Poisson's ratio. Figure 3 , Figure 4 As can be seen, the wave impedance / Poisson's ratio obtained by traditional inversion generally matches the true wave impedance / Poisson's ratio, but uncertainties still exist at the boundaries. In addition to a similar overall trend to the true wave impedance / Poisson's ratio, the wave impedance / Poisson's ratio obtained by diffusion inversion provides a higher resolution description of most sand body boundaries. This demonstrates the reliability and potential of the diffusion inversion method in thin-layer resolution.

[0029] For those skilled in the art, the above description is exemplary and not restrictive. The invention can be implemented in various other forms without departing from the example or spirit of the specification, and thus those skilled in the art can make various obvious equivalent modifications to the invention.

Claims

1. A method for seismic wave impedance diffusion inversion, the method comprising the following steps: Step 1: Select certain wells as seed wells for diffusion inversion. At these well locations, interpret and calibrate the wave impedance attribute curves and Poisson's ratio attribute curves of thin interbedded layers. The original well locations have high-resolution logging curves. Based on the logging curves, interpret the sandstone and mudstone layers. After layering, calculate the mean value of each layer curve, thus forming the initial square wave type calibration layering curve. Step 2: Well Location Inversion; The synthetic trace generated using the calibrated square wave curve is compared with the seismic trace. The interface position and mean value of the square wave curve are adjusted to make the synthetic trace closer to the seismic trace. The adjusted square wave curve is the seed solution for this well location. The process of obtaining the seed solution is as follows: Using a global search algorithm, based on the original square wave layering, the layer thickness, wave impedance attribute curves, and Poisson's ratio attribute curves are appropriately perturbed to obtain hundreds of reflection coefficient curves. Then, a synthetic record is generated based on these results to find the optimal synthetic solution with the smallest difference from the specific seismic record around the selected well location. The final optimal synthetic result corresponds to the new optimal square wave curve layering, the adjusted value, and the well location adjustment, which is the seed solution for the well location. The expression for finding the minimum difference between the synthetic record and the seismic record is as follows: In the formula: To find the sign of the minimum value of the expression; It is the symbol for the second norm in mathematics; Seed solution for reflection coefficient; For the extracted seismic wavelet; The reflection coefficient before the disturbance; For earthquake records; The damping factor is the weight of the two L2 norms in the minimum difference calculation expression; Step 3: Diffusion inversion from well outwards; First, using the seed solution at the well location as a reference, select any adjacent well seismic trace. At this point, the seed solution is considered as R′ in the above expression. Similarly, use the above expression to obtain a new seed solution R for any adjacent well seismic trace location. This expands the original high-resolution seed solution to the seismic trace locations around the well. In this way, throughout the diffusion process, the solution for each new seismic trace location references the seed solution at the adjacent location, maintaining the same number of layers, thus diffusing to the entire seismic data volume. In addition, a time window or layer is added in the vertical direction of the seismic trace to constrain the propagation range, which is the reservoir thickness range.

Images