Derivative-Constrained Reservoir Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current predictive modeling techniques for hydrocarbon reservoirs are complex, requiring large amounts of data and significant computational resources, making them time-consuming and often impractical, especially when data is unavailable or unreliable.
Innovation Solution
The use of derivative-constrained parameterization methods to create compact empirical models of hydrocarbon reservoirs, which involve imposing constraints on model function derivatives to optimize model parameters, reducing data requirements and simplifying the modeling process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complex predictive models are used to characterize hydrocarbon reservoirs, then model accuracy and reliability are improved, but data requirements and computational resources increase significantly
Solution Approach 1:
The patent extracts and utilizes derivative information (rates of change) from available reservoir data to constrain and simplify the empirical model. By focusing on derivative constraints rather than requiring comprehensive datasets, the model achieves reliable predictions with significantly reduced data requirements, directly resolving the contradiction between model accuracy and data quantity needed.
Solution Approach 2:
The patent transforms the modeling approach by changing from direct parameter fitting to derivative-constrained parameterization. This parameter transformation allows the model to capture essential reservoir behavior through constraints on rates of change, achieving high reliability while reducing the quantity of data and computational resources required.
2Reliability
If complex predictive models are used to characterize hydrocarbon reservoirs, then model accuracy is improved, but computational resources and development time increase
Solution Approach 1:
The patent extracts derivative constraints from available reservoir data and uses these to directly parameterize the empirical model. This extraction approach eliminates the need for extensive computational experimentation and iterative model tuning, reducing development time from months to weeks while maintaining high model accuracy through physically-based constraints.
Solution Approach 2:
The patent performs preliminary action by pre-establishing derivative constraints based on reservoir physics and available data before model parameterization. This preliminary constraint setup guides the entire modeling process, eliminating time-consuming iterative adjustments and accelerating model development while ensuring accurate results through physically-based guidance.
3Device complexity
If traditional empirical models are used, then model simplicity is maintained, but they fail to capture complex reservoir behaviors and require extensive data
Solution Approach 1:
The patent introduces derivative constraints as an intermediary mechanism between simple empirical models and complex reservoir behaviors. These derivative constraints act as mediators that translate complex physical relationships into manageable mathematical constraints, allowing simple model structures to accurately capture complex reservoir behaviors without requiring extensive data or complex computations.
Data Source
AI summary
System and method for parameterizing one or more steady-state models each having a plurality of model parameters for mapping model input to model output through a stored representation of an in-situ hydrocarbon reservoir. For each model, training data representing operation of the reservoir is provided including input values and target output values. A next input value(s) and next target output value are received from the training data. The model is parameterized with the input value(s) and target output value, and derivative constraints imposed to constrain relationships between the input value(s) and a resulting model output value, using an optimizer to perform constrained optimization on the parameters to satisfy an objective function subject to the derivative constraints. The receiving and parameterizing are performed iteratively, generating a parameterized model. Multiple models form an aggregate model of the system/process, which may be optimized to satisfy a second objective function subject to operational constraints.