When And Why To Use Streamlines

Streamline simulation is ideally suited for modeling large heterogeneous multi-well systems dominated by convection. While FD methods can account for the convective terms this is the domain where FD methods fail in practice. Conversely, the streamline method is not well suited to complex physics displacements (high compressibility, capillary effects, complicated phase behavior).

There is no one numerical method that can solve the governing equation for all cases efficiently. Depending on which term(s) dominates, different techniques should be used.

If heterogeneity is believed to have a first-order effect of field performance (convective term dominates), multiple fine scale realizations are required. The simulation method of choice is one that is fast, can simulate fine scale models, and can capture effects due to heterogeneity.

The following is a partial list of problems for which streamlines have been found to work particularly well.

• Ranking of Geological Models
  Streamline-based reservoir simulation is ideal for ranking large geological models because of the efficiency in modeling the transport equations regardless of the degree of heterogeneity. For heterogeneous systems, ranking based on incompressible waterflood can be a good indicator of ranking behavior for more complex black-oil displacements.
• Waterflood Optimization
  Because streamlines quantify the connectivity of injectors and producers as a function of reservoir geology, well placement and rates, PVT properties, etc., it is possible to optimally distribute injected volumes to minimize water production while maximizing oil production.
• Fine-Scale Reference Solutions
  Streamline-based reservoir simulation can provide fine-scale reference solutions of the original geological model. This allows to evaluate the effect of different geostatistical algorithms for generating the geological model in the first place and allows to validate upscaling algorithms.
• Novel Data
  By construction, streamlines allow to gather data that is not possible to extract from finite-difference simulators. Streamlines allow to determine well pore volumes and well allocation factors.
• Optimal Infill Drilling
  The optimal location of infill well(s) is dependent on a number of factors, including reservoir heterogeneity, other well locations, and displacement mechanism. In all cases, finding the optimal location requires the minimization of some objective function based on dynamic response data and therefore multiple forward simulations. These forward simulations can be run efficiently using streamlines.
• History Matching Fine Scale Models
  History matching, particularly of water breakthrough and oil production for larger models involves many forward simulations. Additionally, streamlines give information about which areas of the reservoir are "seen" by the various wells and therefore allow to selectively modify geological/flow parameters for history matching purposes.
• Any Voidage Replacement Simulation
  Streamline-based simulation is likely to outperform finite-differences in voidage-replacement type simulations, even if the system is compressible.
• Grid Orientation
  Problems in which finite-differences fail due to grid orientation are good candidates for streamline-based simulation. This is because streamline-based fluid transport does not exhibit grid orientation effects.
• Numerical Diffusion
  As in the case of grid orientation, finite-difference problems that are affected by numerical diffusion are very good candidates for streamline-based simulation. Numerical diffusion can become a problem when simulating strongly heterogeneous systems that might force a small timestep in finite-difference models.
• Sensitivity Studies
  If the displacement is mainly convective dominated, sensitivity studies are best done using streamlines due to the speed of the method.