RESERVOIR SIMULATION
1. Whether the reduction from three-dimensional to two-dimensional numerical algorithms – associated with the migrations of water/oil/gas in large-scale reservoir simulations – towards characterizing water-oil-gas flow system in a petroleum reservoir – would “comfortably/precisely” allow the analysis on the dynamics of water/oil/gas flow system and its associated propagation of water/oil/gas displacement fronts?
If not, to what extent, the reduction into 2D (from 3D) analysis would have an impact on the efficiency of the overall oil recovery process?
Whether the initial saturation distributions of water, oil & gas – associated with such a relatively simpler two-dimensional reservoir simulation would significantly impact the predictions on ‘variations in reservoir rock and fluid properties’?
2. When we solve water/oil/gas flow system through a petroleum reservoir – governed by a set of non-linearly coupled transient PDEs, whether water, oil and gas remain ‘continuous’ ‘at least @ macroscopic-scale’ – ‘throughout the simulation period’?
If not, to what extent, would it impact the assessment of the primary variables that relates velocity fields to the pressure fields through Darcy’s law; and that estimates the saturation distribution through a convective diffusion equation?
3. How exactly the selected numerical algorithms would be able to solve the moving time-dependent oil-water, oil-gas & water-gas interfaces within an oil-water-gas system – with nearly discontinuous water, oil and gas saturation profiles – in the presence of reservoir regions with steep variations in permeability?
4. If the propagation velocity for the moving displacement front remains determined by the water/oil/gas pressure fields; and if these pressure fields remain a strong function of the saturation through the relative permeability functions; then, to what extent, the semi-implicit methods would do justice – in the absence of fully synchronizing both the pressure as well as saturation fields in time?
The importance of geographic dimensionality is function of recovery mechanism (balance of forces, PVT, well positioning,...) and heterogeneity in 3 directions. In some combinations, it is conceivable to perform a dimension reduction with comparatively limited and linear scaling required to get back to extensive physical representation (3D). In other configurations, such an approach is doomed to fail. In many cases, one can represents physical behavior removing a geographic dimension, but the dimension most likely to be removed wo excessive consequences, varies very much. A decision tree approach to the issue is likely to have too many branches to be practical.
- Badges
- Science topic