Seismic Source Separation via Random Delay Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current seismic data acquisition methods face challenges in separating multiple seismic sources activated during a single recording, leading to difficulties in determining the depth of reflectors and interpreting seismic data effectively, especially in areas with intense exploration activity.
Innovation Solution
A method utilizing an inversion-type process to separate seismic recordings containing multiple sources by introducing random time delays between source activations, making interference incoherent and reflections from the same source coherent, allowing for the separation of sources through a numerical inversion process that utilizes sweeps, start times, and coherence of reflection events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple seismic sources are activated simultaneously to reduce survey time and cost, then productivity increases, but the ability to separate and interpret individual source reflections deteriorates
Solution Approach 1:
The patent applies segmentation by dividing the blended seismic signal into individual source components through iterative separation. The method segments the composite recording by identifying and isolating reflections from each source activation, allowing individual depth determination for each reflector while maintaining the efficiency benefits of simultaneous source acquisition.
Solution Approach 2:
The patent employs an inversion approach by working backwards from the blended signal to separate source contributions. The iterative separation process inverts the mixing problem by using the known source waveforms and timing information to reconstruct individual source reflections from the composite recording, thereby resolving the depth ambiguity.
2Loss of time
If multiple seismic sources are activated simultaneously to reduce survey time, then loss of time decreases, but the complexity of data processing and source separation increases
Solution Approach 1:
The patent implements feedback through its iterative separation process. Each iteration uses the previously separated source components as feedback to improve the separation of remaining sources. The method continuously refines the separation by comparing the reconstructed blended signal with the actual recording and adjusting the individual source contributions accordingly, thereby managing processing complexity systematically.
Solution Approach 2:
The patent applies preliminary action by pre-processing the blended signal to identify source-specific characteristics before full separation. The method performs initial sorting of reflections based on known source timing and waveform characteristics, creating a structured framework that simplifies subsequent iterative separation and reduces overall processing complexity.
3Reliability
If random time delays are introduced between source activations to make interference incoherent, then the ability to separate sources improves, but the precision of timing measurements may be affected
Solution Approach 1:
The patent applies parameter changes by systematically varying the time delay parameter between source activations. The method optimizes the random delay values to achieve incoherent interference patterns that facilitate separation while maintaining sufficient timing precision. By controlling the statistical properties of the random delays rather than using fixed values, the method balances separation reliability with timing accuracy.
Data Source
Figure 1
Figure 2
Figure 3~6
AI summary
This is a method of separating simultaneous sources that uses an inversion-type approach. Each source will preferably activated at a random time with respect to the others. These random delays tend to make the interference between sources incoherent while the reflections create coherent events within a series of shots. The shot separation is performed via a numerical inversion process that utilizes the sweeps for each shot, the start times of each shot, and the coherence of reflection events between nearby shots. Implementation of this method will allow seismic surveys to be acquired faster and cheaper.