Reduced-Order Reservoir Models for Faster Forecast Workflows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional hydrocarbon reservoir workflows require significant computational resources and time, delaying decision-making and increasing costs due to the use of high-dimensional parameter space data models.
Innovation Solution
Implementing model-order reduction techniques, history matching techniques, and optimization techniques to generate workflows with a reduced parameter space, using a flow simulator and algorithms like Bayesian optimization to enhance computational efficiency while maintaining accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional high-dimensional parameter space data models are used, then forecast accuracy is maintained, but computational cost and execution time increase significantly
Solution Approach 1:
The patent extracts and identifies a subset of critical input variables from the high-dimensional parameter space using sensitivity analysis and correlation techniques. This extraction process separates the most influential parameters that drive forecast accuracy from the less important ones, allowing the system to maintain accuracy while reducing computational burden by focusing only on the essential variables.
Solution Approach 2:
The patent creates a simplified copy or proxy model that replicates the behavior of the complex high-dimensional model using fewer parameters. This reduced-order model is trained on data from the full model and then used for actual forecasting, providing a computationally efficient alternative that maintains acceptable accuracy for reservoir management decisions.
2Measurement precision
If complex workflows are used, then forecast accuracy is improved, but computational resources and execution time increase
Solution Approach 1:
The patent segments the complex workflow into distinct modular components: data preprocessing, sensitivity analysis, variable selection, model training, and forecasting. Each module performs a specific function and can be independently optimized or replaced. This segmentation reduces overall complexity while maintaining the analytical depth needed for accurate forecasts.
Solution Approach 2:
The patent implements dynamic variable selection where the set of input variables is not fixed but adapts based on the specific reservoir being analyzed. The system dynamically identifies which parameters are most relevant for each case, allowing the workflow complexity to adjust to the actual needs of each forecasting task rather than always using the full complex model.
3Measurement precision
If high-dimensional parameter space models are used, then accurate predictions are achieved, but decision-making speed decreases
Solution Approach 1:
The patent performs preliminary sensitivity analysis and variable importance assessment before the actual forecasting process. This preliminary action identifies and pre-selects the most influential parameters, so that when forecasting is needed, the system already has a reduced set of variables ready, significantly speeding up the decision-making process without sacrificing accuracy.
Solution Approach 2:
The patent changes the parameter representation from high-dimensional raw data to a reduced set of transformed parameters or principal components that capture the essential variability. This parameter transformation maintains the information needed for accurate predictions while reducing the dimensionality, enabling faster computation and quicker decision-making.
Data Source
AI summary
An apparatus for generating forecasts from a high-dimensional parameter data space comprising a reservoir model, a model order reduction module, and an assisted history matching module. The reservoir model having input variables, output variables, and an algorithmic model. The input variables, output variables, and the algorithmic model are generated by a flow simulator module and from a formation and reservoir properties database and a field production database. The model order reduction module generates a subset of the original or transformed input variables. The assisted history matching module adjust values of the output variables based on a difference between the at least one of the output variables and dynamic field production data to improve model accuracy.