Stoneley Dispersion Modeling via Depth Window Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to efficiently model and characterize Stoneley wave dispersion in wells accessing hydrocarbon reservoirs, which is crucial for identifying fractures and rock properties affecting hydrocarbon recovery.
Innovation Solution
A method involving the acquisition of acoustic datasets along wells, determination of depth windows, and the use of dispersion models and nearest neighbor searches to initialize and refine dispersion curves, ultimately characterizing rock properties near the well.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional dispersion modeling methods are used for all depth windows, then measurement precision of rock properties is improved, but computational cost and processing time increase significantly
Solution Approach 1:
The well depth range is divided into multiple depth windows, with a first subset processed using computationally intensive dispersion modeling and a second subset processed using faster nearest-neighbor interpolation. This segmentation allows high-precision modeling only where necessary while reducing overall computational burden.
Solution Approach 2:
Dispersion curves are pre-computed for the first subset of depth windows using accurate dispersion modeling. These pre-computed results serve as reference data for initializing and refining curves in the second subset, eliminating the need to perform full dispersion modeling for every depth window.
2Measurement precision
If traditional dispersion modeling is applied to all depth windows, then rock property characterization is improved, but device complexity and computational resources required increase
Solution Approach 1:
Instead of performing full dispersion modeling for all depth windows, the system copies and adapts dispersion curves from the first subset to the second subset using nearest-neighbor search. This copying approach with subsequent refinement using slowness-frequency pairs maintains accuracy while reducing computational complexity.
Solution Approach 2:
Slowness-frequency pairs serve as an intermediary between the pre-computed dispersion curves in the first subset and the final dispersion curves for the second subset. This intermediary representation enables efficient curve initialization and refinement without requiring full dispersion modeling at each step.
3Loss of information
If comprehensive dispersion analysis is performed across all depth windows, then rock property insights are improved, but processing efficiency decreases
Solution Approach 1:
Full dispersion modeling is performed only for a partial subset of depth windows (the first subset) where it is most needed. For the remaining depth windows (second subset), a refined nearest-neighbor approach with slowness-frequency pair updates is used, which is less computationally intensive but still provides accurate rock property characterization.
Data Source
Figure 1A~1B
Figure 2A~2B
Figure 3
AI summary
Systems and methods for modeling dispersion curves are disclosed. The method includes obtaining an acoustic dataset along a well (20) that accesses a hydrocarbon reservoir (40). The method further includes determining a set of depth windows (120a-f) along the well (20) and determining a first subset of dispersion curves for a first subset of depth windows (120a,d,f) using a dispersion model. The method still further includes initializing a second subset of dispersion curves for a second subset of depth windows (120b,c,e) using a nearest neighbor search of the first subset of dispersion curves. The method still further includes determining slowness-frequency pairs (110) for the second subset of depth windows (120b,c,e) using the acoustic dataset and updating the second subset of dispersion curves using a recursive scanning method. The method still further includes characterizing rock properties near the well (20) based, at least in part, on the first subset of dispersion curves and the second subset of dispersion curves.