Speaker
Description
Enzyme-induced calcite precipitation (EICP) is a bio-cementation technique widely used for soil stabilization, hydraulic control, and groundwater management. By inducing calcite precipitation within pore spaces, EICP modifies the pore structure of porous media and, consequently, alters their flow and transport behavior. Establishing clear links between pore-scale structural evolution and transport dynamics is therefore essential for understanding and optimizing this process. Non-invasive imaging techniques, such as X-ray computed microtomography (µ-XRCT), provide a powerful means to investigate these changes. However, meaningful interpretation of time-lapse µ-XRCT data requires consistent, efficient, and reproducible processing and analysis workflows.
This work presents an analysis of time-lapse µ-XRCT scans acquired during an EICP experiment conducted on a quartz sand packing (15 mm in diameter and 30 mm in height) housed in an X-ray transparent sample holder at the representative elementary volume (REV) scale. The experiment was conducted using a constant flow injection from the sample bottom, while the macroscopic measurement of permeability was also performed. µ-XRCT scans were acquired at predefined pauses throughout the EICP experiment. Each scan consisted of two sequential acquisitions at different stage heights to capture the full sample volume. All scans were performed with an X-ray source voltage of 130 kV, a current of 61 µA, an exposure time of 2.5 s, and 1800 projections over a full 360° rotation, yielding a nominal pixel size of 8 µm.
Following reconstruction, a semi-automated image processing workflow was applied using a combination of open-source and academically available software tools, including FIJI, Dragonfly, and Python libraries. Pore-space segmentation was performed using Otsu’s thresholding method, followed by a fully automated Python-based analysis workflow. This workflow quantified the temporal development of key pore-scale and REV-scale metrics during EICP, including 3D porosity distribution, pore size distribution, average pore diameter, tortuosity, and surface-to-volume ratios.
The image analysis results revealed preferential calcite precipitation near the injection inlet for the adopted EICP injection protocol. Analysis of changes in the pore size distributions in the near-inlet region indicated calcite precipitation occurring in both narrow and large pore spaces. Transport-related changes were inferred from tortuosity trends obtained from REVs distributed along the sample height, which suggested an overall tortuosity increase (aligning well with the sample’s overall permeability impairment measured during the experiment), with the strongest impact near the inlet. In addition, analysis of 3D porosity loss across the sample revealed a distinct connected pathway of calcite precipitation extending from inlet to outlet, providing insight into the development of heterogeneous flow paths during the experiment.
Overall, the proposed computationally efficient workflows, implemented using widely accessible image processing and analysis tools, enable robust processing and analysis of time-lapse µ-XRCT datasets while minimizing user-dependent decisions and ensuring reproducible comparisons across time. Although demonstrated for EICP experiments, the approach is readily adaptable to other time-evolving porous media systems studied using time-lapse µ-XRCT.
Acknowledgment:
This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Project Number 327154368 – SFB 1313.
| Country | Germany |
|---|---|
| Acceptance of the Terms & Conditions | Click here to agree |
