Automatic Field Modeling Calibration via Iterative Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The uncertainty in geological parameters such as rock permeability and basal heat flows, especially over geological times, poses challenges in accurately modeling and simulating field operations, as current methods rely on manual calibration with limited precision.
Innovation Solution
A computer-implemented method and system for automatic calibration of field operations, which iteratively adjusts model realizations based on current measurements, generating likelihood values to select and refine input values, thereby improving the accuracy of modeling results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual calibration methods are used to adjust geological parameters, then the process is simple to implement, but the measurement precision and reliability of modeling results deteriorate due to limited precision
Solution Approach 1:
The patent implements an iterative feedback mechanism where modeling results are continuously compared against actual field measurements, and parameter values are adjusted based on the discrepancies. The system automatically updates parameter values in subsequent iterations using the feedback from measurement comparisons, thereby improving modeling precision through continuous refinement rather than single-pass manual calibration.
Solution Approach 2:
The calibration system performs self-adjustment by automatically modifying parameter values based on measurement feedback without requiring manual intervention. The computer automatically selects which parameters to adjust, determines the adjustment magnitude, and updates the model realization, enabling the system to self-optimize its modeling accuracy through iterative processing.
2Measurement precision
If iterative model calibration is performed to improve accuracy, then the precision of field operation modeling improves, but the time required for processing increases
Solution Approach 1:
The system performs partial calibration by selectively adjusting only certain parameter values rather than all parameters in each iteration. The computer determines which specific parameters require adjustment based on the measurement discrepancies, performing calibrated updates on a subset of parameters to reduce computational burden while maintaining modeling accuracy.
Solution Approach 2:
The system generates multiple model realizations with different parameter sets in advance before the actual calibration process begins. These pre-generated realizations serve as a foundation for rapid iterative calibration, allowing the system to quickly evaluate different parameter combinations and converge to optimal values without extensive real-time computation.
3Reliability
If multiple parameter values are adjusted to improve model accuracy, then the reliability of modeling results improves, but the complexity of managing parameter variations increases
Solution Approach 1:
The patent segments the calibration process by treating each parameter value as an independent adjustable element. The computer selectively identifies and adjusts specific parameter values (such as rock permeability, thermal conductivity, or porosity) independently based on which parameters show discrepancies between model predictions and actual measurements, rather than adjusting all parameters simultaneously. This segmented approach manages complexity by focusing adjustments on only the necessary parameters.
4Measurement precision
If real-time data integration is implemented to continuously update models, then the precision of predictions improves, but the computational resources and processing complexity increase
Solution Approach 1:
The system integrates real-time field measurement data into the calibration process through a feedback mechanism. As new measurements become available during field operations, the computer automatically incorporates this data into the model calibration, comparing actual measurements against model predictions and adjusting parameter values accordingly. This continuous feedback loop enables real-time model updating without requiring complex manual data integration procedures.
Data Source
Figure 1.1
Figure 1.2
Figure 2
AI summary
Automatic calibration for modeling a field includes performing model calibration iterations based on measurements of the field. Each model iteration includes obtaining, during the field operation, a current measurement of the field, generating likelihood values corresponding to model realizations of the field, where each likelihood value is generated by at least comparing the current measurement of the field to a modeling result of a corresponding model realization, selecting, based on the likelihood values, at least one selected model realization from the model realizations, generating, by at least adjusting a first input value of the at least one selected model realization, an adjusted model realization of the field based on the at least one selected model realization, and adding the adjusted model realization to the model realizations. Automatic calibration for modeling a field further includes generating a calibrated modeling result of the field based on the adjusted model realization.