Search

International audience; Spectral induced polarization (SIP) has shown potential to detect subsurface reactive transport processes. However, previous laboratory experiments could not associate the SIP response with the temporal development and spatial distribution of reactive transport processes in detail due to the opaque nature of porous media. We developed a novel experimental setup that consists of a millimeter scale 2D transparent porous medium and electrodes for SIP measurements to visually observe reactive transport processes down to the pore scale while performing SIP measurements (i.e. a 2D millifluidic setup). We used this setup to investigate the SIP response of calcite precipitation (CaCO₃). The image analysis and SIP measurements highlight that the SIP response does not necessarily represent the total mass of calcite precipitation in the measurement region. In addition, we numerically simulated the electric field distribution based on the observed calcite precipitation. The results suggested that the heterogeneous distribution of calcite precipitation within the investigated volume had an important influence on the SIP response. Given the heterogeneous and localized nature of many reactive processes, these findings clearly demonstrate the importance of investigating sub-resolution reactive processes for SIP measurements. The proposed novel experimental setup was shown to be suitable for this purpose.

SUMMARY Induced calcite precipitation is used in geotechnics to modify the mechanical and hydrological properties of the underground. Laboratory experiments have shown that spectral induced polarization (SIP) measurements can detect calcite precipitation. However, the results of previous studies investigating the SIP response of calcite precipitation were not fully consistent. This study aims to investigate how the SIP response of calcite depends on solute concentration to explain the differences in SIP response observed in previous studies. A four-phase experiment with SIP measurements on a column filled with sand was performed. In phase I, calcite precipitation was generated for a period of 12 d by co-injecting Na2CO3 and CaCl2 solutions through two different ports. This resulted in a well-defined calcite precipitation front, which was associated with an increase in the imaginary part of the conductivity ($\sigma ^{\prime\prime}$). In phase II, diluted solutions were injected into the column. This resulted in a clear decrease in $\sigma ^{\prime\prime}$. In phase III, the injection of the two solutions was stopped while calcite precipitation continued and solute concentrations in the mixing zone decreased. Again, this decreased $\sigma ^{\prime\prime}$. Finally, the injection rate of the Na2CO3 solution was reduced relative to that of the CaCl2 solution in phase IV. This resulted in a shift of the mixing zone away from the calcite precipitation front established in phase I and an associated decrease of $\sigma ^{\prime\prime}$. These results imply that the SIP response of calcite is highly sensitive to the solute concentration near the precipitates, which may explain previously reported conflicting results.