Publications

The main challenges in hydrocarbon industry are to build reservoir models that can be used to obtain reliable results of spatial distribution. When data coverage is limited, for some mature field case, the well data number is often high, on the other hand seismic data provides additional constraints on the spatial distribution for lithology mapping purposes. It can be a powerful combination to get the desired sand distribution model. Seismic data on its own does not provide the solution and yet has poor vertical resolution. This paper will explain how to combine geological and geophysical data with spatial and statistical concepts to build sand distribution models. Sandstone lithology is favorable rock where hydrocarbon can be trapped. The models are also honoring the spatial correlation but still put spatial heterogeneity into account. This research was conducted at the Niger Delta, Offshore Nigeria. Firstly, discrete lithology log was interpreted using 32-wells by calculating the Vshale cut-off value where the sand and shale lithology is separated. This discrete sand log will be distributed using sequential indicator simulation process. But, the lateral extension of the sand-dominant lithology should be delineated first. Using RMS amplitude and model-based post-stack seismic-inversion, it can show the lateral extension of low impedance and high energy value of RMS amplitude which interpreted have correlation with the sand lithology existence. From the seismic-derived trend, the probable sand lithology distribution direction is along NW-SE direction. This information then will be used as the horizon low-frequency trend for the simulation process. Using the resulted trend, we can guide the sequential indicator simulation to get multiple equiprobable realization sand distribution. Compared with the unconstrained realization, the result with seismic-derived trend give more geologically reasonable model of sand distribution along the NW-SE major direction.