Abstract: Deriving Permeability Transforms for Reservoir Simulation Models: A Practical Example

GIDMAN, JUNE and DENNIS J. FISCHER

Simulation is an important tool for reservoir management. Populating the simulation grid with structure, porosity, permeability, and saturation data is one of its technical challenges. Of these variables permeability is particularly problematic. Currently there are no logging tools that provide generally acceptable measures of permeability. Instead, most often core data are used to derive a correlation-based porosity/permeability transform. While there is commonly some relationship between porosity and permeability, this relationship is rarely sufficient by itself to adequately predict permeability.

We take various approaches to predicting permeability, based on core and log data. Core and log data must first be quality controlled and depth matched. Since porosity is an important component of most transforms, it is essential that porosity derived from log interpretation compares well to porosity from core. We find that permeability predictions based on a combination of regression and model-based techniques most often provide the best results. There is generally some fine tuning to take into account Vshale, fluid content (gas, oil, or water), reservoir lithology, and borehole rugosity. Our assessment of the best transforms is made by comparison of core and log data on depth plots. The calculated data are subsequently input to geostatistical modeling, scale-up, history matching, and reservoir simulation. The transforms are specific to the reservoir studied, but the approaches can be applied to other fields where there is core and log data.

AAPG Search and Discovery Article #90942©1997 AAPG International Conference and Exhibition, Vienna, Austria