Assessing the effect of pore-scale heterogeneity on glass beads treated with microbially induced calcite precipitation

Traditional soil improvement has negative impact in a world that is seeing an extreme demographic growth, with consequent larger land use and need for infrastructure. In recent years, the growing need to find alternative has led to the development of new technologies in the fields of civil and environmental engineering. Among these, Microbially Induced Calcite Precipitation (MICP), represents a technology capable of improving the mechanical characteristics of soils, with a lower environmental impact. The technology is still in an experimental state and extensive research has been conducted to fully understand the processes behind it, from the laboratory to the field scale. Numerous unknowns related to understanding MICP at the microscale are still present today, from treatment homogeneity to CaCO3 deposition mechanism. The present thesis focuses on the experimental study of the effects of pore-scale heterogeneities on MICP, investigating the influence that the soil heterogeneity has on micro and macroscopic outcomes. For this purpose, three granulometries of increasing gradation (uniform, poorly-graded and well-graded) were tested with a two-phase injection of bacteria solution from top and cementation solution from bottom. The results were explored by means of chemical monitoring of the effluents, calcite content calculation and numerical analysis of its distribution along the treated columns. They were then correlated to the changes in permeability and pore size, which are a function of the variability of the granulometries under study. The results were then verified by means of qualitative scanning electron microscopy (SEM) observations. Chemical monitoring revealed a reaction efficiency greater than 50% for all 11 days of treatment, demonstrating decreasing but always effective bacterial activity. The calcite content estimates coming from the ions concentrations proved to be strongly consistent with the empirical measurements, suggesting that chemical monitoring is a highly effective tool for a non-invasive reaction control and for the prediction of the result. Calcite content was estimated between 2.5% and 5.3% by means of the empirical calculation, while the chemical monitoring reported values on average 20% higher. The results clearly brought to light an increasing treatment inhomogeneity for increased granulometry variability, highlighting the great influence of permeability variation on the pathways followed by the reactants inside the solid matrix. The presence of preferential or obligatory flow paths in the most heterogeneous granulometries was observed, in which the flow slowdowns and filtering effects caused non-uniform CaCO3 precipitation. The SEM observations confirmed most of the numerical results, starting from the preferential distribution of the crystals on the beads of intermediate size, up to the explanation of the limited decrease in permeability following the treatment. Finally, a UCS test was performed on a sub-sample of uniform granulometry, demonstrating an excellent improvement in mechanical properties. The specimen went from zero strength prior to treatment to a UCS of 250 kPa following bio-cementation, reporting a failure mechanism of brittle type. This study provides a greater understanding of the interactions between soil pore-scale heterogeneities and MICP, representing a starting point for the development of new research that would allow to reach a full understanding of the process at the microscale.