INTEGRATED GEOSTATISTICAL SEISMIC INVERSION TO PREDICT A THIN SANDSTONE RESERVOIR UNDER FAN DELTA SEDIMENTARY ENVIRONMENT: A CASE STUDY IN THE JUNGGAR BASIN

Well drilling proved that the Permian's rough grained sandstone reservoir in the centre of Junggar basin was not fully controlled by structure. This type of reservoir is traditionally difficult to identify on conventional seismic data because of the lack of an effective workflow to detect lithology and fluid using real logs and seismic data. In this case study of PetroChina's Xinjiang oilfield, China we present an effective workflow for describing the spatial distribution of sandstone reservoirs and for the feasibility of fluid detection, explaining the reasons why the wells drilled within the higher area of the structure produce water while the other wells located within the lower area produce oil. Forward models derived from petrophysics and rock physics defined the resolution of deterministic inversion, the AVO response of the real reservoir sequence, which inversion method and which elastic parameter were most sensitive to lithology and fluid changes under the real geological environment. In addition, they provide a benchmark for seismic data quality. Higher-resolution reservoir characterization of geostatistical inversion gave the sandstone spatial distribution, the reservoir discontinuity and the horizontal thickness. This integrated workflow guarantees that the final geological map is meaningful not only in terms of elastic parameters, but also in terms of geology, petrophysics, geophysics and engineering.