Multiple-Point Geostatistical Modeling of Braided Channel Reservoir with Constraints by 3D Seismic Data: A Case Study of M Block in Venezuela

Integrating seismic information into multi-point statistical geological modeling can improve the cross well prediction function of the model. Taking a braided river sedimentary oil-bearing block in Orinoco heavy oil belt, Venezuela, as an example, combined with the geological characteristics of braided river reservoir in this area, and using the multi-point statistical modeling method integrated well with seismic data, is this paper it studies the facies calibration of seismic wave impedance, the selection of sand-body probability generation curve, training image analysis, impact ratio between well and seismic data, and their role in multi-point statistical modeling of braided river reservoir. Then, combined with the sedimentological characteristic of braided river reservoir, the modeling results of microfacies spatial distribution such as channel bar, river channel and flood plain in the study area are analyzed. Finally, the uncertainty of different reservoir modeling results is analyzed. The research shows that the multi-point statistical modeling method integrated well and seismic data can better reduce the uncertainty of modeling results in areas with sparse well pattern. Through the probability generation curve of sandstone, the wave impedance data is transformed into the spatial probability distribution of seismic facies. In this way, the relationship between seismic data and its geological significance is effectively established. Compared with modeling only with well-logging data, the method integrated well seismic data modeling result is more reasonable for inter well microfacies prediction, and the continuity of predicted river channel and channel bar is better displayed.