A 3D Geological Modeling Method for Grain-size Lithofacies of Tight Sandstone Gas Reservoirs: A Case Study of Xujiahe Formation Gas Reservoir in Sichuan Basin, China

Abstract

Tight sandstone gas reservoir has strong heterogeneity, and it is difficult to control the change of internal storage and permeability properties by simple sedimentary facies. It is necessary to explore a geological modeling method that can more accurately characterize the distribution of high-quality reservoirs. Taking the Xujiahe Formation gas reservoir in Xinchang area, Sichuan Basin, China as an example, this paper introduces the modeling method of three orders: sand-mudstone facies, sedimentary facies and grain-size lithofacies, so as to realize the spatial characterization of high-quality tight sandstone reservoirs with strong heterogeneity. In this technical process, the sand-mudstone model is first established. Based on the sand and mudstone model, four sedimentary types (or sedimentary facies) of distributary channel, channel edge, interdistributary bay and mouth bar are further divided. The sedimentary facies model is constructed by multi-point geostatistical modeling method. The quantitative relationship between sedimentary facies and grain-size lithofacies is linked by using argillaceous content as a “bridge”. The spatial distribution probability of grain-size lithofacies is constrained by the neural network clustering of argillaceous content and natural gamma-ray data in three-dimensional space. By controlling the types, ratios, and boundaries of grain-size lithofacies through sedimentary facies, and combining probabilistic bodies to cooperate with constraints, precise simulation of lithofacies can be achieved. The grain-size lithology lithofacies model established by this method follows the depositional law in space and has more reasonable contact relations between various sand bodies. The anastomosis rate of the model reached 85% with the new drilling test. This paper provides a new modeling idea for quantitative characterization and prediction of high-quality tight sandstone gas reservoirs, and provides a more accurate model basis.