Large-volume FIB-SEM 3D reconstruction: An effective method for characterizing pore space of lacustrine shales

Abstract

Focused ion beam scanning electron microscopy (FIB-SEM) is a commonly used three-dimensional (3D) pore-network reconstruction method for shales due to its unique capability in imaging nano-scale pores. However, it has been found that for pore space of lacustrine shales with strongly heterogeneous pore structures, the conventional FIB-SEM 3D models usually with dimensions of 10 μm × 10 μm × 10 μm cannot adequately characterize the pore structures as the representative element volume required is much larger than the FIB models. Here, we propose to utilize large volume FIB-SEM (LV-FIB-SEM) 3D models to resolve this challenge. The LV-FIB-SEM model has a significant enhancement in the model size compared with the commonly used conventional FIB-SEM models and a much higher spatial resolution than non-synchrotron nano X-ray CT models for similar imaging sample sizes. With 75 μm × 65 μm × 60 μm as predesigned reconsruction size, after image processing two LV-FIB-SEM 3D models with sizes of 73.56 μm × 38.13 μm × 52.59 μm and 74.01 μm × 43.05 μm × 42.00 μm and model resolution of 30 nm were reconstructed and quantitatively analyzed. When use the conventional FIB-SEM models of 10 μm × 10 μm × 10 μm, the relative deviations between the porosities derived from 100 stochastic models and the average porosity for the two samples studied are −41.13% ∼ +87.31% and −51.66% ∼ +56.05%, respectively, indicating that such small models are not representative of the actual pore structure of the shales investigated. When the model sizes have been increased by 96 times volumetrically, the probabilities of matching average porosities for the two samples increase from 13% to 86% and from 12% to 100%, respectively. This research demonstrates that the upsizing of the FIB-SEM models enables an effective improvement on the representativeness of shale pore structures characterized. It is recommended that LV-FIB-SEM 3D reconstruction be employed to study pore space of lacustrine shales with strongly heterogeneous pore structures, which would enable a more accurate characterization and evaluation of reservoirs for shale oil exploration and development.