NMR Viscosity Modeling via Radial Basis Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current NMR logging technologies face challenges in accurately modeling subterranean fluid viscosity due to limitations in interpreting NMR data, particularly in differentiating between live and dead oil and handling noise in measurements, which affects the accuracy of viscosity predictions.
Innovation Solution
The implementation of a radial basis function (RBF) model combined with principal component analysis (PCA) and the use of apparent hydrogen index (HIapp) to preprocess and normalize T2 distributions from NMR measurements, allowing for more accurate viscosity predictions across varying echo times and reducing the impact of noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional NMR logging methods are used to measure subterranean fluid viscosity, then the measurement process is simple, but the accuracy of viscosity predictions is poor due to inability to differentiate between live and dead oil and sensitivity to noise
Solution Approach 1:
The patent segments the NMR signal analysis by separating live oil and dead oil components through principal component analysis of T2 distributions. This segmentation allows the model to independently characterize each fluid type, improving viscosity prediction accuracy by capturing their distinct relaxation behaviors rather than treating them as a single homogeneous signal.
Solution Approach 2:
The patent introduces an intermediary processing layer between raw NMR data and viscosity prediction. This intermediary layer uses principal component analysis to transform T2 distributions into standardized representations, and employs noise filtering techniques to eliminate measurement artifacts before the final viscosity calculation, thereby improving prediction reliability.
2Loss of information
If NMR measurements are taken with varying echo times to improve fluid characterization, then the amount of information obtained increases, but the complexity of processing and normalizing the data increases
Solution Approach 1:
The patent applies preliminary normalization and principal component analysis to T2 distributions before viscosity prediction. By pre-processing the data to establish consistent reference frames across different echo times, the model eliminates the need for complex real-time adjustments during measurement, reducing processing complexity while preserving all fluid characterization information.
Solution Approach 2:
The patent transforms the T2 distribution data through parameter changes including normalization to reference echo times and conversion to principal component space. These parameter transformations standardize the data representation across varying echo times, allowing the model to integrate information from multiple echo times without proportionally increasing processing complexity.
3Reliability
If noise filtering is applied to NMR measurements to improve accuracy, then the reliability of viscosity predictions improves, but the processing time and computational requirements increase
Solution Approach 1:
The patent extracts and removes noise components from NMR measurements through principal component analysis, which separates signal from noise by identifying and eliminating variance not correlated with fluid properties. This extraction approach selectively removes only the harmful noise elements while preserving the meaningful signal, improving reliability without requiring excessive processing time.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the predictive accuracy of subterranean fluid viscosity, improving the model's performance by considering the shape of T2 distributions and accounting for differences between live and dead oil, while also mitigating the effects of noise, resulting in viscosity estimates that are generally within one order of magnitude of measured values.
Implementation Method 1
nuclear magnetic resonance (NMR) tools have been used to explore the subsurface based on the magnetic interactions with subsurface material
Data Source
AI summary
Systems, methods, and software for modeling subterranean formation viscosity are described. In some aspects, a method of training a subterranean fluid viscosity model based on NMR data includes accessing multiple relaxation-time distributions generated from NMR measurements of a fluid, normalizing each relaxation-time distribution to a common normalizing value, computing parameters for a plurality of weighted radial basis functions from the normalized relaxation-time distributions, and producing a subterranean fluid viscosity model that includes the weighted radial basis functions and the computed parameters.