Reservoir Model Calibration via Sound Sequence Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional reservoir surveillance techniques face challenges in accurately monitoring and calibrating fluid and rock properties in multi-million-cell hydrocarbon reservoirs, particularly in giant reservoirs with poor seismic data quality, leading to inefficiencies in data analysis and model refinement.
Innovation Solution
A computer-implemented method that converts numerical predictions of reservoir variables into sound sequences for comparison with actual monitoring measurements, allowing for interactive calibration of reservoir models and optimizing data acquisition based on predicted changes, using a petro-elastic model to evaluate the suitability of monitoring technologies and determine optimal measurement intervals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional visualization and data-mining techniques are used for multi-million-cell reservoir simulation, then the simulation can be performed with existing computer power, but the data analysis becomes a significant bottleneck and cannot promptly interact with vast amounts of data
Solution Approach 1:
The patent replaces conventional visual inspection and data-mining approaches with an automated event monitor system that uses syntax rules to automatically detect and analyze reservoir events. This substitution of manual/mechanical analysis with an automated computational system eliminates the data analysis bottleneck while maintaining the ability to process multi-million-cell simulation data efficiently.
2Measurement precision
If direct measurement techniques such as 4D seismic monitoring are applied to giant reservoirs, then dynamic changes in the reservoir can be detected, but the changes may take many years to be observable with precision and require good seismic signal quality which is not always available
Solution Approach 1:
The patent performs preliminary action by using reservoir simulation to predict expected dynamic changes in reservoir properties before actual monitoring begins. These predictions are converted into sound sequences that can be compared with actual monitoring data, allowing for earlier detection and calibration without waiting for slow direct measurement techniques to show changes over many years.
Solution Approach 2:
The patent introduces sound sequence conversion as an intermediary between numerical reservoir predictions and actual monitoring measurements. This intermediary transformation allows for more sensitive and faster comparison of predicted versus actual reservoir behavior, enabling earlier detection of changes without requiring good seismic signal quality.
3Reliability
If extensive reservoir surveillance equipment is deployed to monitor giant reservoirs, then comprehensive data can be collected, but unnecessary investment in monitoring hardware may occur and data analysis becomes complex
Solution Approach 1:
The patent applies partial action by using reservoir simulation to predict which dynamic changes are most significant and converting only those predictions into sound sequences for monitoring comparison. This selective approach provides reliable monitoring of critical reservoir behaviors without deploying extensive surveillance equipment for all possible parameters, thereby reducing hardware investment and system complexity while maintaining monitoring accuracy.
Data Source
AI summary
A computer-based system generates digital and audio responses to changes in fluid and rock properties of a producing hydrocarbon reservoir for surveillance analysis. The system calibrates observed changes against directly-measured field data in order to optimize the reservoir model. The changes may include, for example, stress changes in rock, impedance changes in rock, and fluid density changes.