A new adaptive approach is proposed to the creation of geological and hydrodynamic models of the fields and reservoirs with long production history, taking into account the limited amount and large uncertainty of the available initial data. In the adaptive geological model, the number of layers of its grid does not exceed the number of layers identified by the results of detailed correlation, while all model parameters are constructed depending on seismic data. The grid of the adaptive hydrodynamic model completely coincides with the geological grid without any upscaling. In the adaptive hydrodynamic model, the system of differential equations describing the filtration process is solved according to the rules of the theory of percolation and cellular automata. Since at the entrance of the adaptive hydrodynamic model, there are not the flow rates of fluids, but the volumes of oil and water produced, as well as the volumes of injection, such model is always history matched. Calculation of the adaptive hydrodynamic model allows generating the distribution of current oil reserves and reservoir pressure in the field or the reservoir at any time. Using the adaptive hydrodynamic model, it is also possible to predict the state of further development of the field or the reservoir, using the iterative fuzzy-logical method to determine the oil decline rates first, and then to solve the system of equations for the interference of the wells. The proposed adaptive geological-hydrodynamic model is not simplified, it is complex and requires practically the same estimated time as the deterministic model. However, the adaptive model is easier for users, since it does not require any manual work and is always adjusted to the fact, so it turns out to be cheaper and faster than its deterministic analogue.
- Cellular automata
- Forecast of oil production
- Fuzzy logic
- Geological and hydrodynamic model
- Theory of percolation