Nested Fourier-DeepONet for 3D geological carbon sequestration (GCS)

The data and code for the paper J. E. Lee, M. Zhu, Z. Xi, K. Wang, Y. O. Yuan, & L. Lu. Efficient and generalizable nested Fourier-DeepONet for three-dimensional geological carbon sequestration. Engineering Applications of Computational Fluid Mechanics, 18 (1), 2024.

Data

The full dataset for Nested Fourier-DeepONet is available on OneDrive. Download and put all files into datasets folder.

Steps to generate the full dataset for Nested Fourier-DeepONet:

• Step 1: Download raw data
• Step 2: Run file_config.sh to convert .npy file into .pt files
• Step 3: Create a new folder datasets_nested_fno under the main folder and put the files generated in step 2 into the folder datasets_nested_fno
• Step 4: Run data_generation.py to covert .pt files into our .npz files

Code

Fourier-DeepONet training

Pressure buildup models:

• Level 0: Run dP0_global_training.py
• Level 1-4: Adjust the level of interest in line 12 and run dP1-4_LGR_training.py

Gas saturation models:

• Level 1-4: Adjust the level of interest in line 12 and run SG1-4_LGR_training.py

Output generation

Raw data files are required.

Pressure buildup:

• Level 0: Run dP0_create_output.py
• Level 1-4: Adjust the level of interest in line 12 and run dP1-4_create_output.py

Gas saturation:

• Level 1-4: Adjust the level of interest in line 12 and run SG1-4_create_output.py

Evaluation

Outputs saved in .npz files are required to run the following evaluation code.

Pressure buildup:

• Run dP_evaluation_code.py

Gas Saturation:

• Run SG_evaluation_code.py

Cite this work

If you use this data or code for academic research, you are encouraged to cite the following paper:

@article{lee2024efficient,
  author  = {Jonathan E. Lee, Min Zhu, Ziqiao Xi, Kun Wang, Yanhua O. Yuan and Lu Lu},
  title   = {Efficient and generalizable nested Fourier-DeepONet for three-dimensional geological carbon sequestration},
  journal = {Engineering Applications of Computational Fluid Mechanics},
  volume  = {18},
  number  = {1},
  pages   = {2435457},
  year    = {2024},
  doi     = {https://doi.org/10.1080/19942060.2024.2435457}
}

Questions

To get help on how to use the data or code, simply open an issue in the GitHub "Issues" section.