Capacitive conductivity logging and electrical stratigraphy in a high-resistivity aquifer, Boise Hydrogeophysical Research Site

Abstract

We tested a prototype capacitive-conductivity borehole tool in a shallow, unconfined aquifer with coarse, unconsolidated sediments and very-low-conductivity water at the Boise Hydrogeophysical Research Site ͑BHRS͒. Examining such a high-resistivity system provides a good test for the capacitive-conductivity tool because the conventional induction-conductivity tool ͑known to have limited effectiveness in high-resistivity systems͒ did not generate expressive well logs at the BHRS. The capacitive-conductivity tool demonstrated highly repeatable, low-noise behavior but poor correlation with the induction tool in the lower-conductivity portions of the stratigraphy where the induction tool was relatively unresponsive. Singular spectrum analysis of capacitive-conductivity logs reveals similar vertical-length scales of structures to porosity logs at the BHRS. Also, major stratigraphic units identified with porosity logs are evident in the capacitive-conductivity logs. However, a previously unrecognized subdivision in the upper portion of one of the major stratigraphic units can be identified consistently as a relatively low-conductivity body ͑i.e., an electrostratigraphic unit͒ between the overlying stratigraphic unit and the relatively high-conductivity lower portion -despite similar porosity and lithology in adjacent units. The high repeatability and resolution and the wide dynamic range of the capacitive-conductivity tool are demonstrated here to extend to high-resistivity, unconsolidated sedimentary aquifer environments.

The hydrogeologic information in cross-borehole complex conductivity data from an unconsolidated conglomeratic sedimentary aquifer

“…A contrast of only one order of magnitude in real conductivity is apparent; this variation is dominated by the difference between cobble-dominated units (units 1-4) and the sand unit (unit 5), the latter showing higher real conductivity that is consistent with the observed high ∅ of these sediments (Figure 3). The capacitive probe logs reported in Mwenifumbo et al (2009) as conductivity show a similar weak contrast, in their case between 0.5 and 3 mS∕m (−3.3 to −2.5 range in log 10 σ 0 , expressed in S∕m). More subtle variation in real conductivity exists throughout the cobble-dominated units, although a distinct contrast is seen at approximately 838 and 844 masl (e.g., at C5 in Figure 8a).…”

“…Cluster analysis was carried out using either five or six clusters; these values were chosen because application of multiple analyses with identical values for all other parameters showed the greatest consistency with five clusters, slightly reduced consistency with six clusters, and poor consistency with four or seven clusters. Furthermore, analyses with five or six clusters allowed comparisons to be made with previous interpretations of BHRS stratigraphy based on ∅, lithology, and CR (Barrash and Clemo, 2002;Barrash and Reboulet, 2004;Mwenifumbo et al, 2009). Each identified cluster could include any nonzero proportion of the total number of voxels in the grid and could have irregular and discontinuous boundaries.…”

“…Figure 12. Cluster analysis results for well C4: (a) real conductivity profile from inversion, (b) imaginary conductivity profile from inversion, (c) cluster boundaries from five-cluster analysis, (d) cluster boundaries from six-cluster analysis, and (e) unit boundaries (Barrash and Clemo, 2002;Mwenifumbo et al, 2009). …”

“…The original five-unit stratigraphy was extended following the capacitive probe resistivity (CR) well logging outlined in Mwenifumbo et al (2009). With this technique, the electrical resistivity close to the well is measured using two pairs of coaxial metal cylinders forming a transmitter capacitor and a receiver capacitor, operating at a frequency of approximately 1 MHz.…”

“…Subtle changes in d 60½matrix are discernible at the boundaries between units 2A and 2B, between units 2B and 3, and between units 3 and 4. A high-resolution distribution of the six stratigraphic layers in the central region of the BHRS was developed following Barrash and Clemo (2002) and Mwenifumbo et al (2009) to provide a "ground-truth," not only primarily of porosity distributions but also of electrical parameters, against which results of other geophysical surveys can be compared. The distribution is based on a kriged geostatistical model that includes all stratigraphic unit contact elevations in the wells.…”