Study on Local Brittleness of Rock Based on Multiple Linear Regression Method: Case Study of Shahejie Formation

Abstract

Rock brittleness index is an important parameter that affects the volume fracturing effect of an unconventional reservoir. The higher the brittleness index, the larger the communication area of fracture network formed under volume fracturing. However, rock brittleness has strong heterogeneity, which seriously affects the propagation of volume fracture network. Most existing rock brittleness characterization methods start from macro brittleness and cannot analyze the influence of heterogeneity on reservoir brittleness. The mineral composition analysis method can characterize the heterogeneity of a reservoir by analyzing the influence of mineral composition on brittleness, but this is limited to qualitative analysis on different minerals, resulting in large errors. In this study, the relationship between macro mechanical properties and micro mineral composition of rocks was analyzed using multiple linear regression analysis and a large number of experimental core parameters on the Shahejie Formation to understand the quantitative characterization of rock brittleness distribution in a heterogeneous reservoir. The results showed that quartz, plagioclase, analcite, siderite, and pyrite were high brittle minerals while clay minerals were low brittle minerals. This method provides a new focus for the analysis of brittle distribution of heterogeneous reservoirs and an additional method to study rock fracture law.