The integration of magnifications: A novel approach to obtain representative information about the pore space of mudrocks from SEM images

Impact of Drying and Effective Stresses on the Pore Space and Microstructure of Mudrocks

“…Although the oven‐dried SEM imaging porosity can be estimated at a higher resolution, the analysis was restricted to the 35 nm‐limit to allow for comparative analysis between oven‐dried and frozen‐hydrated samples. Investigating the fabric of a sister sample to the one studied in this paper at 1 MPa (Table ), Deirieh, Casey, et al () measured the oven‐dried SEM imaging porosity at a resolution of 11 nm and found it to be 0.17, an increase of ~20% compared to the oven‐dried SEM imaging porosity of 0.14 measured in this investigation at a resolution of 35 nm (Table ). The oven‐dried SEM imaging porosity at a resolution of 11 nm is still significantly less (representing ~60%) than the oven‐dried gravimetric porosity of 0.28.…”

“…Because of experimental limitations of cryoSEM, we did not attempt to determine the porosity at smaller scales. Narrow confidence intervals (95% probability level) indicate that the number of scales and number of images at each scale are sufficient to represent the pore space at the microscale (Deirieh, Casey, et al, ). The SEM imaging porosity of oven‐dried samples decreased from 0.16 to 0.10 as stress increased from 0.2 to 10 MPa, while the cryoSEM imaging porosity of frozen‐hydrated samples (1 and 6 MPa) ranged from 0.11 to 0.12 (Figure and Table ).…”

“…SEM imaging porosity was determined as the area fraction of pores identified in digital SEM images, using the integration of magnifications (IOM) method developed by Deirieh, Casey, et al () to account for a very wide distribution of pore sizes such as those present in mudrocks. The IOM method accounts for the issue that for any given SEM image, the largest pore area that can be accurately determined has to be smaller than the total area imaged (in other words, the pore has to fit entirely within the image).…”

“…Manual thresholding is user dependent with an expected variability of ±10% (Houben, ). The resulting images were subsequently analyzed by a MATLAB® code that implements the IOM method described above (Deirieh, Casey, et al, ).…”