Multiple–Point Geostatistical Lithofacies Simulation of Fluvial Sand–Rich Depositional Environment: A Case Study From Zubair Formation/South Rumaila Oil Field

Abstract

SummaryModeling accurate geological structure is an important step in reservoir simulation. The multiple-point facies geostatistics (MPS) has been successfully used as an efficient approach to capture complex geological structures, in comparison with the sequential indicator simulation. Specifically, MPS creates the most realistic depositional environments because it mainly depends on using training images that reflect actual geological architecture, rather than variogram. Therefore, the MPS algorithm was adopted to model the 3D lithofacies distribution of the fluvial sand-rich depositional environment of the main pay in South Rumaila Oil Field in Iraq.The fluvial depositional system was created through a 2D user-defined training image that was sampled to create a surface map. Then, a neural-networks algorithm was adopted to train the surface map to create the discrete template, which then was used with the 3D grid construction for pattern generation of 3D facies distribution. The resulting training image represents a numerical geological model that contains 3D information about the reservoir's geological heterogeneity and structure. The facies pattern was then considered for 3D lithofacies modeling conditioning to upscaled discrete facies distribution for 19 wells in the reservoir.The discrete lithofacies distributions of sand, shaly sand, and shale, obtained from core-analysis data, have been modeled as a function of well-log interpretations in one well and then were predicted for all 19 wells distributed around the reservoir crest. The training-image body and channels were pointed to the three aforementioned facies to create the 3D facies system. The resulting integrated model, through MPS, has much more reasonable facies architecture than indicator simulation because it is clearly noticeable how sand channels are distributed with continuous direction toward the sea shoreface. To validate our model and to capture the true depositional environment, many realizations were generated on the basis of the leave-one-out cross-validation technique. The sand-channels continuity was achieved through spatial posterior probability of lithofacies distribution that was incorporated in the MPS approach as a 3D trend model. However, including the geometrical trend models of thickness or depth has no effect on the channels continuity. The spatial posterior probability trends were modeled from the predicted probability distributions of lithofacies, which have been created by the kernel support vector machine (KSVM) for all 19 wells in the formation.The resulting MPS model for the main pay of South Rumaila oil field corresponds to the description of the fluvial depositional environment described in the literature. This reflects how MPS is efficient to reconstruct complex geological reservoirs. In addition, it preserves the reservoir heterogeneity and connectivity of flow paths, and it provides a realistic reservoir description for petrophysical-property modeling and reservoir flow simulation.