Seismic Data Internal Multiple Removal via Pseudo-Depth Aggregation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The computational cost of inverse scattering series (ISS) for three-dimensional seismic data processing is prohibitive, making it difficult to effectively remove internal multiples, which lead to artifacts and incorrect interpretation in seismic data analysis.

Innovation Solution

A two-step prediction approach is implemented, transforming seismic data to a pseudo depth domain, aggregating and then inversely transforming it back to the acquisition domain to generate three-dimensional predictions of internal multiples, reducing computational complexity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If inverse scattering series is applied to three-dimensional seismic data processing, then internal multiples can be predicted, but computational cost becomes prohibitive

Engineering Contradiction:
Improveinternal multiple prediction accuracyVSAvoidcomputational cost
Core Design Contradiction:
Measurement precisionVSUse of energy by stationary object

Solution Approach 1:

The patent divides the three-dimensional seismic data processing into multiple two-dimensional sub-problems. By segmenting the 3D volume into 2D slices and processing each slice independently through inverse scattering series, the computational complexity is reduced from O(N³) to O(N²), making internal multiple prediction feasible for large-scale seismic datasets while maintaining prediction accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the three-dimensional internal multiple prediction problem into a series of two-dimensional problems by introducing a dimensional reduction approach. This dimensionality change allows the application of inverse scattering series in 2D space, significantly reducing computational requirements while still capturing the essential 3D characteristics of internal multiples through stacking and integration of 2D results.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If three-dimensional inverse scattering series is performed, then complete internal multiple prediction is achieved, but processing time becomes excessive

Engineering Contradiction:
Improveprediction completenessVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the 3D seismic volume into multiple 2D sub-volumes or slices that can be processed independently and in parallel. This segmentation enables distributed computing approaches where different 2D slices are processed simultaneously across multiple computational cores or nodes, dramatically reducing total processing time while maintaining complete coverage of the 3D survey area.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent processes 2D slices with appropriate overlap and integration to achieve complete 3D coverage. By performing partial 2D inverse scattering series on overlapping sub-volumes and combining results through proper stacking and interpolation, the method achieves comprehensive 3D internal multiple prediction without the full computational burden of processing the entire 3D volume at once.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4624999A1Methods and computing systems for internal multiple removal from seismic data
Publication Date: 2025.10.01 SERVICES PETROLIERS SCHLUMBERGER SA
  • EP4624999A1 patent drawingFigure 1
  • EP4624999A1 patent drawingFigure 2
  • EP4624999A1 patent drawingFigure 3

AI summary

A method implements internal multiple removal from seismic data. The method involves transforming seismic data, including a set of internal multiples, from an acquisition domain to a pseudo depth domain to generate pseudo depth data. The method further involves aggregating the pseudo depth data to generate aggregated data. The method further involves inverse transforming the aggregated data from the pseudo depth domain to the acquisition domain to generate two dimensional prediction data. The method further involves aggregating the two dimensional prediction data to generate three dimensional prediction data representing a set of internal multiple predictions corresponding to the set of internal multiples.