Waterflood Injector-Producer Pair Modeling for Rapid Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current simulation techniques for oil and gas operations using detailed models are computationally intensive, taking weeks or months to generate enough optimization scenarios, making real-time optimization challenging and impacting long-term production performance due to unscheduled events.
Innovation Solution
Implementing machine learning models to generate performance indicators from reduced data, which are then processed by a pattern flood management application, allowing for rapid simulation results while maintaining accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detailed numerical simulation models are used for waterflood optimization, then measurement precision and reliability are improved, but productivity and response time deteriorate due to computational intensity taking weeks or months
Solution Approach 1:
The patent creates simplified surrogate models that replicate the behavior of complex numerical simulation models. These surrogate models are trained on data from detailed simulations and can predict waterflood performance with acceptable accuracy in minutes rather than months, enabling real-time optimization while maintaining essential model fidelity
Solution Approach 2:
The patent divides the reservoir into discrete injector-producer pairs and represents their relationships through separate transfer coefficients. This segmentation allows the complex coupled simulation problem to be decomposed into independent pairwise interactions, dramatically reducing computational complexity while preserving the essential physics of fluid flow between injectors and producers
2Measurement precision
If detailed numerical simulation models are used for waterflood optimization, then measurement precision is improved, but loss of time increases due to computational intensity taking weeks or months
Solution Approach 1:
The patent performs preliminary simulations to train surrogate models offline before real-time optimization is needed. The surrogate models are pre-trained on comprehensive simulation data covering various operating conditions, so that during actual optimization operations, predictions can be made instantly without repeating the full numerical simulation process
Solution Approach 2:
The patent creates simplified surrogate models that replicate the behavior of complex numerical simulation models. These surrogate models are trained on data from detailed simulations and can predict waterflood performance with acceptable accuracy in minutes rather than months, enabling real-time optimization while maintaining essential model fidelity
3Reliability
If complex numerical simulation-based studies are used for waterflood management planning, then measurement precision and reliability are improved, but device complexity increases making timely modifications difficult
Solution Approach 1:
The patent replaces complex numerical simulation models with simplified surrogate models that capture the essential relationships between injectors and producers. These surrogate models maintain sufficient reliability for strategic planning while being computationally lightweight enough to allow rapid modification when operational conditions change
Solution Approach 2:
The patent divides the reservoir into discrete injector-producer pairs and represents their relationships through separate transfer coefficients. This segmentation allows the complex coupled simulation problem to be decomposed into independent pairwise interactions, dramatically reducing computational complexity while preserving the essential physics of fluid flow between injectors and producers
4Loss of information
If unscheduled field and well events occur during detailed simulation-based operations, then loss of information increases about subsurface conditions, but productivity decreases due to inability to react in timely manner
Solution Approach 1:
The patent implements a monitoring system that continuously tracks performance indicators of injector-producer pairs and compares them against predicted values from surrogate models. When deviations indicate unscheduled events, the system provides feedback that triggers rapid re-optimization using the lightweight surrogate models, enabling timely responses to subsurface condition changes that would be impossible with traditional detailed simulation approaches
Data Source
AI summary
A method including receiving injector-producer pair parameters for injectors and producers in a target underground region. The injectors and the producers may be characterized as injector-producer pairs. The method also includes converting the injector-producer pair parameters into coefficients stored in a data structure. The coefficients represent estimates of connection strengths between the injectors and producers. The method also includes generating, from the coefficients, a performance indicator that represents an operational relationship between a corresponding injector and a corresponding producer in an injector-producer pair. The method also includes transmitting the performance indicator to a pattern flood management application.