Uncertainty Analysis of an Integrating Two Structural 3D Sthochastic Geological Models of a Siliciclastic Reservoir: Potiguar Basin, Northeast of Brazil

Abstract

Abstract The studied field is located in the Brazilian Continental Shelf, the Potiguar Basin. These reservoirs are Eo–Cretaceous sandstones, which correspond to the gradual passage from the continental to marine depositional environments. The workflow used was chosen in order to reflect the concern in trying a more effective representativeness of the structural complexities of the studied field. It consists in building several three-dimensional stochastic reservoir models and in recalculate the STOOIP (Stock Tank Originally Oil In Place), based on the integration and the interpretation of available data. Two structural geological models were built with distinctive geophysical interpretations based on a seismic reference surface. These tectonic differences can modify the stratigraphic distribution of the reservoirs and change drastically the fluid flow on the petrophysical system of hydrocarbons distributions on the reservoirs. The uncertainty analysis was built and the parameters that impacted the most the STOOIP were varied under each considered structural model. These parameters were: the range of the porosity values and the cut-off of oil. A risk curve of 520 simulations was developed, that identifies which models corresponds to the respective percentiles and obtain the respective realization to simulate on the flow the many probabilistic sets. We create the Hydrocarbon Pore Volume maps – HPhiSo, for each percentile of each structural model. The percentiles P50 of these maps of the two models were superimposed, and it was defined a set of planned wells to the reservoir drainage. This permits to relate the biggest hydrocarbon column with the highest structural map, reducing the risk. The geological uncertainties analysis results from numeric numbers varying the STOOIP until 37%, while oil was considered in all blocks. Eight realizations of each 3D geological model will be submitted to flow simulation and the planned wells will be ranked according to their productions curves.