Data acquisition and processing of multi-frequency oscillatory hydraulic tomography in a granular aquifer
Résumé
Transport of solutes in aquifers is controlled by the heterogeneous spatial distribution of hydraulic properties, but the characterization of aquifer heterogeneity is quite challenging with conventional methods. Hydraulic tomography (HT) was developed to define the heterogeneous distribution of hydraulic conductivity (K) and specific storage (Ss). HT involves the emission of a series of hydraulic head perturbations in a stressed well and the recording of this signal at several levels in the stressed and observation wells. All recorded hydraulic head responses are simultaneously analyzed through numerical inversion, which provides the spatial distribution of hydraulic properties at a scale relevant for local site investigations.
This communication reports on a tomographic experiment carried out in a heterogeneous and highly anisotropic granular aquifer at the Saint-Lambert research site near Quebec City, Canada. This site has already been the object of detailed characterizations with multiple hydraulic methods: pumping tests, packer slug tests, flowmeter profiles, vertical interference tests, and slug test tomography. A relatively new approach named oscillatory hydraulic tomography (OHT) was tested, in which multi-frequency oscillatory head perturbations are induced in an interval isolated by packers of the stressed well by a submerged rod that is electronically controlled by a winch system. Hydraulic responses are measured in the stressed intervals and in multiple intervals of an observation well.
This study was primarily aimed at testing, first on an operational level, if the OHT signal could be generated in the stressed well and propagated to the observation well in a highly anisotropic granular aquifer. Second, the study developed a rigorous workflow for the treatment of the measured hydraulic heads. Third, in terms of characterization efficacy, the study aimed to determine if multiple controlled frequencies would allow the assessment of K spatial distribution.
Results show that the field experiment provided clear measured hydraulic responses that could be used to obtain the 2D distribution of hydraulic properties from the inversion of OHT measurements. Comparison was made of inversion results using a single oscillatory frequency and multiple frequencies. Under conditions of realistic field measurement noise and uncertainty, it will be valuable in future work to compare the imaging capabilities of oscillatory hydraulic tomography against other tomographic methods. Further investigation is also needed to examine the information content of oscillatory hydraulic tomographic data for characterizing K and Ss heterogeneities through a sensitivity and resolution analysis. This study demonstrates the practical potential for the implementation of OHT experiments in relatively low permeability and highly anisotropic granular aquifers.
