AI Variogram Modeling for Faster Reservoir Parameter Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The labor-intensive process of variogram modeling for spatial and vertical data in earth modeling is time-consuming, especially when using global optimization algorithms like Genetic Algorithms, which are not optimized for time-critical processes such as live earth modeling workflows.
Innovation Solution
A method involving a deep learning system that uses synthetic data to predict initial variogram parameters, which are then calibrated by a local optimization algorithm and used as input for a Genetic Algorithm, significantly reducing the processing time required for variogram fitting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If global optimization algorithms like Genetic Algorithm are used to find best-fit variogram parameters, then measurement precision is improved, but loss of time increases significantly
Solution Approach 1:
The patent applies preliminary action by training a machine learning model in advance using synthetic well data and variogram parameters. The trained model can then quickly predict variogram parameters for new well data without requiring time-consuming Genetic Algorithm iterations, thus resolving the contradiction between accuracy and computation time.
Solution Approach 2:
The patent uses synthetic well data that copies the characteristics of real well data to train the machine learning model. This synthetic data approach allows the model to learn from numerous examples without requiring actual field data for each prediction, enabling fast and accurate parameter estimation.
2Manufacturing precision
If manual variogram modeling is performed for each zone and facies, then manufacturing precision is improved, but productivity decreases
Solution Approach 1:
The patent implements self-service by automating the variogram modeling process through a machine learning system. The system automatically processes well data, trains models, and generates variogram parameters without requiring manual intervention for each zone and facies, thereby maintaining precision while dramatically improving productivity.
Solution Approach 2:
The patent creates a universal machine learning model that can handle multiple zones and facies types with a single system. The model is trained on diverse synthetic data representing different geological conditions, enabling it to generalize and provide accurate variogram parameters across various reservoir zones without requiring separate manual modeling for each case.
Data Source
AI summary
A method for variogram modeling is disclosed. The method includes obtaining a synthetic well data and a well data for facies or petrophysical properties of interest in a targeted reservoir zone, training machine learning models using the synthetic well data as inputs and outputting a plurality of final variogram parameters predicted by using a plurality of machine learning models for the facies or petrophysical properties of interest in the targeted reservoir zone, wherein the well data for the facies or petrophysical properties of interest in the targeted reservoir zone is used as input.