Reservoir Simulation Oscillation Suppression via Dynamic Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current numerical reservoir simulators face computational challenges due to oscillations caused by flow reversals in low drawdown wells, particularly in thermal simulations, which require user-supplied parameters to limit time step sizes, making the process computationally expensive and often prohibitive.
Innovation Solution
A method and system that analyze simulation variable values from grid cells to generate a scaling factor, adjusting these values to control intermediate result deviations between computation iterations, thereby reducing oscillations and optimizing simulation results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If user supplied parameters are used to limit time step size, then oscillations due to flow reversals are mitigated, but computational cost increases and becomes prohibitive
Solution Approach 1:
The patent dynamically adjusts the scaling factor parameter during simulation based on detected oscillation conditions, rather than using fixed user-supplied parameters. This adaptive parameter change allows the system to maintain simulation stability only when oscillations are present, avoiding the continuous computational overhead of restrictive time step limitations.
Solution Approach 2:
The system implements feedback by monitoring simulation variable values between computation iterations to detect oscillations. When oscillations are detected, the scaling factor is adjusted to suppress them. This closed-loop feedback mechanism enables automatic stabilization without requiring pre-defined restrictive parameters from users.
2Measurement precision
If time step size is restricted to mitigate oscillations, then computational accuracy is improved, but the number of computation iterations increases
Solution Approach 1:
Instead of applying restrictive time step limitations throughout the entire simulation, the patent applies partial action by only adjusting the scaling factor when oscillations are actually detected in specific grid cells. This selective approach maintains accuracy where needed while avoiding unnecessary computational overhead in stable regions.
Solution Approach 2:
The system transitions from static time step restrictions to dynamic scaling factor adjustment. The scaling factor is computed and applied adaptively during the simulation based on real-time detection of oscillation conditions, allowing the system to respond only when and where accuracy issues arise, rather than imposing continuous restrictions.
3Reliability
If scaling factor is used to control intermediate result deviation, then oscillations are reduced, but computational complexity increases
Solution Approach 1:
The patent extracts the oscillation suppression function into a separate, modular scaling factor computation that operates independently from the main simulation algorithm. By taking out this specific function, the core simulation code remains unchanged while adding a relatively simple oscillation detection and correction layer that computes scaling factors based on monitored variable deviations.
Data Source
AI summary
From a simulation engine simulating the field operation, simulation variable values corresponding to grid cells are obtained. The grid cells represent at least a portion of the field for simulating the field operation. The simulation variable values are analyzed to generate a scaling factor and adjusted according to the scaling factor to generate adjusted simulation variable values. The scaling factor controls intermediate result deviation between computation iterations of simulating the field operation. Using the simulation engine, a simulation result is generated based at least on the adjusted simulation variable values. A field operation is performed based on the simulation result.