NMR Logging Data Inversion for Formation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current NMR logging tools face challenges in combining data from different depths of investigation (DOIs) to maintain precision and accuracy, particularly due to varying fluid distributions and poor signal-to-noise ratios, which can lead to inconsistent datasets and erroneous interpretations.
Innovation Solution
A partially constrained 4-dimensional inversion method is employed to combine NMR measurements from different DOIs, analyzing shared and distinct properties to compute common and distinct distributions, respectively, thereby improving precision while maintaining independence of fluid distributions across DOIs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data from all DOIs are combined and a single inversion is performed, then measurement precision is improved, but fluid distribution accuracy deteriorates due to varying fluid distributions across DOIs
Solution Approach 1:
The patent segments the fluid distribution into distinct components: bound fluid (shared across all DOIs) and free fluid (distinct at each DOI). This segmentation allows the inversion to treat different fluid types differently, combining data for bound fluid to improve precision while maintaining DOI-specific independence for free fluid to preserve accuracy.
Solution Approach 2:
The patent applies local quality by assigning different constraints to different parts of the solution space. Bound fluid parameters are constrained to be consistent across all DOIs (local consistency), while free fluid parameters are allowed to vary by DOI (local independence). This enables precision improvement through data combination where appropriate while maintaining accuracy where fluid distributions differ.
2Reliability
If measurements are acquired to allow independent inversion at each DOI, then fluid distribution accuracy is maintained, but measurement precision deteriorates particularly at deeper DOIs
Solution Approach 1:
The patent merges data from multiple DOIs in a selective manner. For bound fluid components, data from all DOIs are combined in a single inversion to improve measurement precision through increased signal averaging. For free fluid components, data remain effectively separate to maintain accuracy. This selective merging resolves the precision-accuracy tradeoff.
Solution Approach 2:
The patent creates a composite inversion model that combines elements of both fully independent and fully combined approaches. The model uses a composite structure where bound fluid inversion uses combined data from all DOIs (improving precision) while free fluid inversion uses DOI-specific data (maintaining accuracy), effectively creating a hybrid solution that leverages the strengths of both approaches.
3Productivity
If a single inversion is performed combining all DOIs, then the number of measurements and total time are reduced, but inconsistent datasets and erroneous interpretations result
Solution Approach 1:
The patent introduces dynamic constraints into the inversion process. Rather than using fixed constraints for all parameters, the methodology dynamically adjusts constraints based on the physical nature of different fluid types. Bound fluid parameters receive strong coupling constraints across DOIs, while free fluid parameters receive weak or no coupling constraints, allowing the inversion to adapt to the varying reliability of data at different DOIs.
Data Source
AI summary
A method to determine formation properties using two or more data sets in which the solutions corresponding to the data sets represent shared and distinct formation properties. The method analyzes the data sets and computes distributions for the shared and distinct formation properties from which the formation properties are determined.