This repository contains the code for the research work titled "Optimising Waterflooding Strategies for Enhanced Oil Recovery with Physics-Informed Neural Networks," which has been submitted to the SPE Journal.
The code in this repository is structured to facilitate the training and analysis of Physics-Informed Neural Networks (PINNs) for optimizing waterflooding strategies in oil recovery. It includes scripts for both 2D and 3D cases, as well as a script for optimizing water injection rates using a genetic algorithm.
-
Train_PINNs_2D.py: Script for training the 2D case of the PINN model. -
analysis_2D.py: Used for analyzing the model's results for the 2D case. -
Train_PINNs_3D.py: Script for training the 3D case of the PINN model. -
analysis_3D.py: Used for analyzing the model's results for the 3D case. -
optimisation.py: Script to optimize the water injection rate using a genetic algorithm.
To get started with this project:
- Clone the repository to your local machine.
- Ensure you have the necessary dependencies installed.
- Run the training scripts for either the 2D or 3D case as per your requirement.
- Analyze the results using the corresponding analysis scripts.
- For optimization, execute
optimisation.py.
MIT License
For any queries regarding the code or the research, please reach out to h.meng94@outlook.com.