Geostatistics provides powerful solutions to create geological models which integrate all the available sources of information. Well data, seismic data, sedimentological interpretations and basic reservoir engineering results can be incorporated to improve the accuracy and realism of geological models.
Detailed fluid behavior analyses in EOR studies can be based on such geological models. Geostatistical geological models help improve reservoir production forecasts and reduce costs by facilitating Production History Match.
Petroleum reservoirs production forecasts are based on flow simulations which are using a Geological Static Model as main input. The accuracy and realism of the geological model are critical parameters when optimizing local recovery with EOR methods, in areas with remaining hydrocarbons.
Geostatistical techniques are used for building the structural model and for populating the 3D grid with properties. They account for the accuracy of each data source.
Geostatistics enhance many components of a geological model, leading to significantly improved reservoir modeling results:
STRUCTURAL MODELING
By removing acquisition and processing artifacts, geostatistical filtering techniques lead to accurate velocity models, which improve time-to-depth conversion of geological horizons.
Horizons mapping can be enhanced by using advanced estimation techniques which integrate geological trends and uncertain data, each source of information being accounted for according to its accuracy.
FACIES MODELING
Geological modeling is usually done hierarchically by facies, each facies being characterized by a specific and quite homogeneous petrophysical properties distribution. Geostatistics offers many facies simulation methods which all honor well data and account for geological trends, by mean of facies proportions 3D models. In particular Plurigaussian technique (PGS) allows reproducing sedimentological environments.
PETROPHYSICAL PROPERTY MODELING
With geostatistical mapping techniques, it is easy to populate the geological grid with porosity and permeability inside each facies, honoring well data and geological trends. Domain analysis methods allow detecting possible border effect, i.e. smooth petrophysical properties transitions across facies borders. Such border effects can be accounted for by advanced geostatistical mapping algorithms.
Post-processing and iterative modeling methods allow integrating basic reservoir engineering results, such as information about connection between wells, leading to geological models consistent with actual flow behavior
Detailed fluid behavior analyses in EOR studies can be based on such geological models. They also allow a significant cost reduction by facilitating Production History Match duration.