MCI Logging Hybrid Inversion for Formation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-component induction (MCI) logging methods for subterranean formations face accuracy issues due to radial one-dimensional inversion algorithms and are affected by strong shoulder-bed and horn effects, especially in complex borehole environments, leading to degraded results.
Innovation Solution
A hybrid enhancement of radial one-dimensional (R1D) and vertical one-dimensional (V1D) inversion processing is implemented, combined with adaptive low-pass filtering and Software Focusing (SWF) processing, to improve the accuracy of formation log data and reduce horn effects, using a multi-step borehole correction scheme for MCI tools.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radial one-dimensional inversion algorithm is used for MCI logging, then processing speed is maintained, but measurement precision deteriorates due to shoulder-bed and horn effects
Solution Approach 1:
The inversion processing is divided into two distinct segments: radial one-dimensional (R1D) inversion and vertical one-dimensional (V1D) inversion. Each segment handles specific aspects of the data processing, with R1D dealing with conventional resistivity and V1D addressing anisotropy and dip calculations. This segmentation allows each algorithm to be optimized for its specific purpose while avoiding the computational burden of a single complex algorithm.
Solution Approach 2:
The patent transitions from traditional single-dimension radial inversion to a two-dimensional approach by incorporating vertical dimension processing through V1D inversion. This dimensional expansion enables the system to handle complex borehole environments and anisotropic formations more accurately by considering both radial and vertical variations in formation properties.
2Measurement precision
If adaptive low-pass filtering and Software Focusing processing are applied, then measurement precision improves, but loss of time increases due to additional processing steps
Solution Approach 1:
Software Focusing processing is applied as a preliminary step before the main inversion processing. By pre-processing the MCI data to enhance vertical resolution and reduce noise, the subsequent inversion steps work with already-optimized data, reducing the computational burden and processing time required for the main analysis.
Solution Approach 2:
Adaptive low-pass filtering acts as an intermediary process between data acquisition and final inversion. It selectively removes high-frequency noise while preserving important formation signals, thereby improving measurement precision without requiring excessive processing power or time in the final inversion stage.
3Reliability
If multi-step borehole correction scheme is implemented, then reliability of formation parameter determination improves, but device complexity increases
Solution Approach 1:
The borehole correction is implemented as a multi-step process with distinct correction stages. Each step addresses specific error sources or environmental factors separately, allowing for systematic correction of measurement inaccuracies while maintaining clear documentation and control of each correction applied to the data.
Data Source
AI summary
Hybrid inversion processing techniques are implemented that result in improved speed and accuracy of determining formation properties using log data, for example, from an multi-component induction logging tool. Logging data relating to the formation of interest is obtained and used as an input. High frequency noise is then removed from the logging data and bed-boundary determination is performed using the logging data. An adaptive low pass filter is applied to the logging data and the logging data is inverted. The inverted logging data is correct and enhanced by determining one or more weights based on one or more quality indicators. The inverted logging data may then be visually interpreted and used to adjust one or more drilling parameters.