Reservoir characterization by geostatistics: Amarume oil field case.

Geostatistics has been recognized as a powerful tool for the reservoir characterization since it can integrate geological, geophysical, and reservoir engineering data. Generally, the measurements derived from the well data are relatively precise, however, they are located limitedly in the reservoir. On the contrary, seismic data is obtained uniformly in the reservoir area, and therefore, very useful in the reservoir characterization. The relationship between seismic information and reservoir properties is still not clear because of its limitation of frequency band width, low S/N ratio, and the ambiguity of the data. Thus, the relationships may vary with region and with depth.In order to deal with such ambiguous relations, geostatistical techniques, such as cokriging and cosimulation, have been adopted since they are probabilistic approaches. Also geostatistics can integrate well and seismic data in order to estimate the spatial distribution of reservoir properties.A case study of reservoir characterization in the Amarume oil field is presented in this paper by using the geostatistics. The acoustic impedance was derived from the 3-D seismic volume by the inversion method. A relationship between acoustic impedance and lithology can be observed in and around the sand reservoir. In this field, the sand formations show lower impedance, compared with other lithological layers. Applying this relationship, we can infer the spatial distribution of sand formations on the 2-D section tying the two wells. The Sequential Indicator Cosimulation (SICOSIM), which is one of the geostatistical methods, was applied in this analysis. And equiprobable multi-realizations of the sand distribution were estimated. These realizations are consistent with both well log data and seismic data.