Downhole SARA Fraction Prediction Using ML and Optical Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining SARA fractions of reservoir fluids are limited by the need for laboratory analysis, which is costly and time-consuming, and fail to provide real-time data on asphaltene precipitation conditions, leading to potential production disruptions and inefficiencies in oil operations.
Innovation Solution
A machine learning-based system that uses downhole sensors and a Downhole Fluid Analyzer module to predict SARA fractions from real-time fluid composition measurements, enabling accurate estimation of asphaltene behavior and precipitation conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional laboratory SARA analysis is used, then measurement precision is improved, but loss of time and productivity deteriorate
Solution Approach 1:
The patent replaces conventional mechanical laboratory analysis systems with an optical measurement system using a downhole spectrometer. The spectrometer uses optical sensors to measure fluid properties directly in the wellbore, eliminating the need for time-consuming laboratory processing while maintaining measurement accuracy through calibrated optical detection methods.
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between raw optical measurements and SARA fraction calculations. The ML model processes spectrometer data and converts it into accurate SARA fraction predictions, enabling real-time analysis without requiring traditional laboratory infrastructure or manual analysis procedures.
2Measurement precision
If laboratory fluid analysis is performed, then measurement precision is improved, but productivity deteriorates
Solution Approach 1:
The patent performs SARA fraction measurements in advance within the wellbore before production issues occur. By continuously monitoring fluid composition in situ, the system identifies asphaltene precipitation risks early, allowing operators to take preventive actions before production is disrupted, thus maintaining both accuracy and productivity.
Solution Approach 2:
The patent substitutes traditional laboratory-based compositional analysis with an optical measurement system that provides real-time fluid characterization in the wellbore. This replacement enables continuous monitoring and rapid decision-making, significantly improving productivity while maintaining analytical precision through calibrated optical detection.
3Productivity
If downhole measurements are used, then productivity is improved, but device complexity increases
Solution Approach 1:
The patent designs the downhole measurement system to perform multiple functions: the spectrometer measures various fluid properties (SARA fractions, asphaltene content, precipitation conditions) simultaneously using a single optical platform. This multi-functionality reduces overall system complexity compared to having separate specialized instruments for each measurement type.
Solution Approach 2:
The patent uses a machine learning model as a computational intermediary that simplifies the relationship between complex optical measurements and SARA fraction predictions. The ML model handles the complexity of data processing and interpretation, allowing the physical measurement hardware to remain relatively simple while achieving accurate real-time predictions.
4Reliability
If real-time monitoring is implemented, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent implements real-time monitoring with continuous feedback capability, where the spectrometer continuously measures fluid composition and the ML model continuously predicts SARA fractions and precipitation risks. This feedback loop provides ongoing reliability information to operators, enabling real-time adjustments to production parameters to prevent asphaltene precipitation without requiring overly complex intervention systems.
Data Source
AI summary
Systems and methods estimating the SARA fractions of a reservoir fluid. This method uses a machine learning (ML) based model to predict the SARA fractions of a reservoir fluid. The ML models are trained using conventional laboratory data, such as fluid composition from gas chromatography, SARA measurement, Asphaltene onset pressure (AOP) etc. Reservoir fluid can be pumped from a wellbore into a downhole fluid analyzer tool. The downhole fluid analyzer tool can take measurements indicating the presence and levels of certain particles in the fluid. The measurements can be applied to the ML models to estimate SARA levels in the reservoir fluid.