Programming Language: Python 3 (fipy, matplotlib, numpy)
Project Сode: FipyFracSolver.ipynb
Project Full Description (in Russian): Description.pdf
Company: Gazpromneft Technology Partnerships LLC.

Earlier in the project (2021) Complex Predictive Model Transfer the complex model was transferred from Excel format to VBA format. This model does not take into account hydrodynamics, which leads to the need for manual fitting of the model. Existing software does not allow taking into account a key reservoir property (permeability depends on pressure) when calculating production volume, which leads to the need to adapt models by varying reservoir parameters and hydraulic fracture properties arbitrarily - this leads to the risk of making incorrect decisions in terms of multi-stage hydraulic fracturing technology.

The main goal of this project is the introduction of a hydrodynamic model into the forecast complex. Tasks:

• selection of the optimal model solver for performance;
• comparison of model results with the commercial simulator tNavigator;
• forecast of pressure and flow rates;
• implementation of the prototype in Python.

• Selected the optimal solver for the model in terms of computation speed (finite element method with mesh thickening in the crack area, the model is a non-linear diffusion equation in partial derivatives);
• Adapted the solver to the conditions of the problem, developed a Python 3 prototype and made a forecast of the production volume of one well.
• 5% - maximum deviation from the commercial simulator tNavigator.

Input data: description of the model from company experts, operator of a commercial simulator.
Problems: artifacts of model nonlinearity.
Methods: nonlinear differential equations, numerical and analytical methods, Python 3, Jupyter Notebook.

The results of the prototype: the distribution of pressure around the well and the volume of oil production per day.

Project Сode: FipyFracSolver.ipynb
Project Full Description (in Russian): Description.pdf