Dynamic Geo-Model Calibration via Pressure Transient Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the field of geologic modeling, existing methods struggle to accurately predict hydrocarbon-bearing formation properties away from well control points, especially in green fields where limited production data are available, leading to variability in geological model robustness.
Innovation Solution
The RMS Workflow generates dynamically calibrated geo-models by simulating multiple adjusted geo-models based on pressure transient data, identifying the best fit by minimizing the difference between simulated and measured bottomhole pressure changes, and refining geological property values to create a robust and accurate contour map of hydrocarbon-bearing formations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple geological realizations are generated using different variogram, azimuth, seeds, and co-kriging settings, then the model can account for uncertainty in property distribution away from wells, but the variability and robustness of properties away from well control points become difficult to evaluate
Solution Approach 1:
The patent implements an iterative feedback mechanism where simulated pressure transient data from multiple geological realizations is compared against actual field measurements. The comparison results feed back into the model calibration process, allowing automatic adjustment and ranking of realizations based on how well they reproduce observed pressure behavior. This feedback loop systematically evaluates model robustness without requiring manual screening of hundreds of realizations.
Solution Approach 2:
The patent systematically varies key geological modeling parameters including variogram range, azimuth, random seeds, and co-kriging weights to generate diverse geological realizations. By controlling and documenting these parameter changes, the method creates a structured set of realizations that can be efficiently evaluated against pressure transient data to assess property distribution robustness away from wells.
2Loss of information
If pressure transient data is used to test model robustness away from wells, then information from a radius of 1 km or more around tested wells can be utilized, but significant production data is not available on green fields
Solution Approach 1:
The patent performs preliminary calibration of geological models using pressure transient data available from early delineation and development wells before full production data becomes available. By utilizing pressure transient responses and available production rates from initial wells, the method establishes a calibrated baseline model that can be used for field development planning, rather than waiting for extensive production data to accumulate.
Solution Approach 2:
The patent uses pressure transient data as an intermediary to bridge the gap between limited well control points and the broader formation properties. Pressure transient responses serve as a mediator that provides information about formation characteristics at distances up to 1 km from wells, enabling model calibration and validation without requiring dense production data coverage across the entire green field.
3Measurement precision
If properties at well level are made consistent with core and log data, then the model satisfies basic validation criteria, but the algorithm and parameters used to distribute properties between well control points remain highly sensitive
Solution Approach 1:
The patent transitions from static property distribution methods to a dynamic calibration approach where geological model parameters are systematically adjusted based on pressure transient data matching. Instead of relying on fixed algorithms and parameters for property distribution, the method dynamically optimizes variogram parameters, co-kriging weights, and other distribution controls through iterative comparison of simulated versus observed pressure responses, reducing sensitivity to initial parameter choices.
Data Source
AI summary
Techniques to generate dynamically calibrated geo-models green fields are described. A geo-model representing a field on which wells are drilled in a hydrocarbon-bearing formation adjusted to generate multiple adjusted geo-models. Each adjusted geo-model represents a variant of the numerical geo-model. Using each adjusted geo-model, multiple simulated rates of change of bottomhole pressures over time in a well drilled in the hydrocarbon-bearing formation are determined. A measured rate of change of bottomhole pressures over time in the well is compared with the multiple simulated rates of change of bottomhole pressures over time in the well. Based on a result of the comparing, the adjusted geo-model that yielded simulated rates of change of bottomhole pressures that best matched the measured rate of change of bottomhole pressure is identified. A geological property associated with the best-match adjusted geo-model is determined and presented in a geological property contour map of the hydrocarbon-bearing formation.