NMR Fluid Substitution via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Nuclear magnetic resonance (NMR) interpretation models are not applicable for rock formations with multiphase fluids, as the presence of hydrocarbons disturbs T2 distributions, invalidating the assumption of a correlation between NMR distributions and pore size distributions, necessitating a method for accurate fluid substitution.
Innovation Solution
A machine learning-based NMR fluid substitution model is trained to identify hydrocarbon responses in T2 distributions and replace them with water responses, using mineralogy, lithology, total porosity, and water saturation data, allowing for the prediction of T2 distributions of 100% water saturation, which can be applied to oil zones within a reservoir formation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If NMR interpretation models are applied to multiphase fluid-bearing formations, then the analysis can be performed using existing models, but the results are inaccurate because hydrocarbons disturb T2 distributions and invalidate the correlation assumption with pore size distributions
Solution Approach 1:
The patent extracts and removes the hydrocarbon component from the multiphase T2 distribution to isolate the water signal. By identifying and separating the hydrocarbon response portion of the T2 distribution, the method enables accurate pore size distribution analysis based solely on the water signal, which maintains the valid correlation with pore geometries.
Solution Approach 2:
The patent segments the multiphase T2 distribution into distinct hydrocarbon and water components. This segmentation allows the water portion to be analyzed separately using established NMR interpretation models, while the hydrocarbon portion is identified and removed, resolving the contradiction between model applicability and measurement accuracy.
2Measurement precision
If conventional NMR fluid substitution methods are used to replace hydrocarbon responses with water responses, then the T2 distribution can be corrected, but the process requires core and fluid analyses which are time-consuming and costly
Solution Approach 1:
The patent uses T2 distributions from water-bearing zones as copies or proxies for the water component in oil-bearing zones. By assuming similar petrophysical characteristics between water zones and oil zones, the method creates a synthetic water T2 distribution that can be directly applied to correct the multiphase T2 distribution without requiring physical core samples or fluid analyses.
Solution Approach 2:
The patent enables the NMR logging data itself to provide the necessary information for fluid substitution. By using the T2 distributions from water zones within the same formation to correct the oil zone T2 distributions, the method makes the system self-sufficient, eliminating the need for external core and fluid analysis processes.
3Ease of manufacture
If T2 distributions from water zones are used to replace hydrocarbon responses in oil zones, then fluid substitution can be performed, but the accuracy depends on the assumption that petrophysical characteristics are similar between water and oil zones
Solution Approach 1:
The patent applies local quality by selecting water zones that specifically match the petrophysical characteristics of the target oil zone. Rather than using any water zone, the method identifies and uses water zones with similar mineralogy, lithology, and porosity properties, thereby improving the reliability of the fluid substitution while maintaining the simplicity of the approach.
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
Enables accurate determination of rock properties such as permeability and pore size distribution without requiring core and fluid analyses, improving well planning and hydrocarbon recovery operations by generating reliable petrophysical models for multiphase fluid-bearing formations.
Implementation Method 1
Nuclear magnetic resonance (NMR) logging tools have been used in the oil and gas industry to explore the subsurface based on the magnetic interactions with subsurface material. NMR logs, including longitudinal (T1) and transverse (T2) relaxation measurements
Data Source
AI summary
System and methods for nuclear magnetic resonance (NMR) fluid substitution are provided. NMR logging measurements of a reservoir rock formation are acquired. Fluid zones within the reservoir rock formation are identified based on the acquired measurements. The fluid zones include water zones comprising water-saturated rock and at least one oil zone comprising rock saturated with multiphase fluids. Water zones having petrophysical characteristics matching those of the oil zone(s) within the formation are selected. NMR responses to multiphase fluids resulting from a displacement of water by hydrocarbon in the selected water zones are simulated. A synthetic dataset including NMR T2 distributions of multiphase fluids is generated based on the simulation. The synthetic dataset is used to train a machine learning (ML) model to substitute NMR T2 distributions of multiphase fluids with those of water. The trained ML model is applied to the NMR logging measurements acquired for the oil zone(s).