Simultaneous-Source Seismic Data Separation via Iterative Sparse Inversion

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for processing simultaneous-source seismic data face challenges in separating signals from individual sources due to spatial aliasing and incoherency issues, especially when the shot interval is large, leading to inefficiencies in data processing and interpolation.

Innovation Solution

A computer-implemented method and system that utilize simultaneous-source separation modules to process seismic data by treating the composite energy signal as a single entity, applying linear operator transforms to decompose it into signals associated with individual sources, and employing iterative techniques such as sparse inversion and random noise attenuation to separate and interpolate the data effectively.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If simultaneous-source acquisition is used to record multiple seismic shots over the same time interval, then acquisition efficiency is increased, but the recorded signals from individual sources cannot be easily separated

Engineering Contradiction:

Improveacquisition efficiencyVSAvoidsignal separation complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The composite signal containing multiple simultaneous sources is segmented into individual source contributions through iterative separation. The method divides the complex multi-source signal into separate source records by systematically isolating each source's contribution, enabling independent processing of each source while maintaining acquisition efficiency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The separation process is performed as a preliminary step before further data processing. By separating sources upfront in the domain containing data from many shots, the method enables subsequent processing steps to work with clean, source-specific data, avoiding the need for re-separation later in the workflow.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If separation techniques are implemented in a domain containing data from many shots to take advantage of dithering, then source separation is enabled, but sampling becomes dependent on shot interval causing spatial aliasing

Engineering Contradiction:

Improvesource separation accuracyVSAvoidspatial aliasing

Core Design Contradiction:

Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The method transitions from shot-domain processing to a common-image-point (CIP) domain, changing the dimensional framework for analysis. This dimensional transformation allows the separation technique to operate on migrated images rather than raw shot data, enabling better handling of spatial sampling issues and reducing aliasing effects while maintaining separation accuracy.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The common-image-point domain serves as an intermediary representation between the raw simultaneous-source data and the final separated source records. This intermediate domain allows the application of separation techniques that are less sensitive to shot interval constraints, acting as a mediator that preserves separation accuracy while avoiding spatial aliasing.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Ease of operation

If conventional interpolation algorithms are applied to unseparated simultaneous-source data, then data processing is simplified, but incoherency from shot to shot prevents effective interpolation

Engineering Contradiction:

Improveinterpolation simplicityVSAvoidinterpolation accuracy

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The data is segmented into individual source records before interpolation is applied. By separating the simultaneous-source data into distinct source contributions first, each source's data becomes coherent and suitable for conventional interpolation algorithms, maintaining both operational simplicity and interpolation reliability.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Source separation is performed as a preliminary action before interpolation. This sequencing ensures that the data is in the appropriate form (separated source records) before interpolation algorithms are applied, enabling both simple and accurate processing without the incoherency problems that would otherwise prevent effective interpolation.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentEP2823336B1Methods and computing systems for processing data

Publication Date: 2019.02.27 WESTERNGECO LLC

Figure 1
Figure 2~3
Figure 4

AI summary

Computing systems and methods for processing collected data are disclosed. In one embodiment, a method for iteratively separating a simultaneous-source dataset is provided, wherein the simultaneous-source dataset is used as an input dataset for a first iteration of simultaneous-source separation. The input dataset includes a plurality of shots that include data corresponding to a plurality of source activations. The method of iteratively separating the input dataset includes generating simulated simultaneous shots based on shots separated in the input dataset; and forming an output dataset based on the separated simultaneous shots and the simultaneous-source dataset, wherein the output dataset is configured for use as the input dataset for the next iteration of separating the simultaneous-source dataset.