Fractal-based NMR permeability estimation in tight sandstone: A case study of the Jurassic rocks in the Sichuan Basin, China

Permeability is a crucial parameter evaluating tight sandstone reservoir. Previous permeability estimation methods based on nuclear magnetic resonance (NMR) are commonly suitable for well-sorted, highly porous and unconsolidated sandstone, but incurring large errors when applied to tight sandstone. This study introduces the fractal dimension based NMR permeability estimation for tight sandstone. Core permeability and NMR measurements were performed on thirty tight sandstone samples selected from the Jurassic rocks in the Sichuan Basin, China. We proposed two fractal-based NMR permeability models for tight sandstone in the studying area, which are considered as the modifications of the SDR and T-C models, examining the applications on core plug samples and comparing the results with other permeability models. Results show that tight sandstones have wider T2 distribution and weaker correlation between throat size and pore size than unconsolidated/clean sandstones. The proposed coefficient-based optimal fitting method can obtain more accurate fractal dimensions of micropores and macropores than these achieved by subjective and empirical methods and commonly used T2cutoff method. The proposed models introduced in this research have two advantages: (1) the models have high accuracy since the effective pore structure parameters (porosity, FFI/BVI and T2LM) are adopted. The complex pore structure of tight sandstones is considered by using fractal dimensions. (2) the models are easy to implement by following the concise formats of the Timur-Coates model and SDR model. In practice, the models can be applied to predict permeability of tight sandstone in the lab, or in the field requiring NMR T2 distributions as basic data, integrating core permeability to determine fitting parameters in the models.