A PTA pattern recognition tool – pta-learn Python library – is released as an open-access outcome of AutoWell: Automated Well Monitoring and Control collaboration project between NORCE, University of Stavanger (UiS) and Heriot-Watt University (HWU).

pta-learn is designed to automate detection key flow regimes based on PTA data, such as radial flow – and pinpointing the end of well bore storage effects. Developed in collaboration with Anton Shchipanov, Vasily Demyanov and Khafiz Muradov as part of a peer-reviewed research study, pta-learn helps Reservoir Engineers to quickly analyze and interpret large PTA datasets.

The PTA pattern recognition algorithms was published earlier in:

• Feature Extraction and Pattern Recognition in Time-lapse Pressure Transient Responses

  Starikov, V., Shchipanov, A., Demyanov, V. & Muradov, K., Nov 2024, In: Geoenergy Science and Engineering. 242, 213160.

Explore the library on PyPI: https://lnkd.in/dUwjtb76
For an in-depth look at our approach, read our peer-reviewed paper: https://lnkd.in/dmeN7xGW
We’ve also included hands-on Jupyter notebooks in Google Colab to help you get started immediately.
Flow regime feature extraction example:
https://lnkd.in/dynWt2jR
Pattern recognition in time-lapse PTA example:
https://lnkd.in/dfSNq62Z