Interpretation of Hydraulic Tests and Implications Toward the Representative Volume Element for Bedrock Systems

A bedrock site was studied over the past four years. The primary objective of site research was in situ bioremediation. During the course of the site research, seven wells in competent bedrock were constructed to augment the pre-existing site investigation wells. The new wells had multiple hydraulics tests performed in various lengths of discrete intervals of the open well bore. These tests included over 200 slug and 10 pumping tests. One pumping test was a site-wide test that involved wells in overburden, weathered bedrock, and competent bedrock. This wealth of information provided a database to make statistical inferences about bedrock hydraulic parameters. In addition, the analysis of the slug test data takes on new meaning when the data is analyzed by a heterogeneous conceptual model rather than the more common homogeneous models. Lastly, due to the varying size of formation investigated by the different hydraulic tests, it is possible to identify the representative volume element (RVE) - the size of formation over which small changes from average hydraulic characteristics occur. Pumping tests included standard constant rate pumping as well as non-constant pumping. Slug tests used gas pressurization and vacuum for traditional changes in instantaneous head (less than one meter), as well as large displacements on the order of 15 meters. For all data sets, data was analyzed with homogeneous analytical models (Hantush, Cooper, etc.) and a simple radial, two-zone heterogeneous finite-difference numerical model. This model used two zones, radially centered on the stressed well, with different hydraulic properties (usually with transmissivity one or two orders of magnitude higher in the outlying zone compared to the zone containing the stressed well). For the pumping tests and the large displacement slug tests, in addition to the analysis of the data from the stressed well, data from monitoring wells was also analyzed to develop hydraulic parameters. Hydraulic parameters, resulting from the analysis of the following hydraulic tests, were similar: the site-wide pumping test, the interval pumping tests, and the slug test far field (outer zone). Since each of these tests, evidenced by monitoring wells, investigated different sized portions of the bedrock and because the resulting hydraulic parameters were similar, this begins to define the bedrock RVE. More importantly, this RVE was identifiable from slug tests in single wells, for single intervals without the use of nearby monitoring wells. At this site, the RVE was on the order of 10 meters. This RVE from hydraulic tests is compatible with what can be estimated from fracture density observed in cores, outcrops, and geophysical logs. The interrelation between hydraulic testing of well bore intervals and testing the entire open well bore may in fact overlap in the description of the RVE. If one considers that an open borehole is a vertical zone in which to study hydraulic parameters, and that this zone is on the order of the RVE, then open well bore hydraulic tests may yield RVE-scale information. Arithmetic and geometric sums of well bore interval transmissivity compared favorably to the open borehole transmissivity, which supports this conclusion. The probability distributions representing the slug test hydraulic conductivity from homogeneous and heterogeneous methods will be presented and discussed. The heterogeneous, far field hydraulic conductivity probability distribution differed from the homogeneous distribution in median, mode, mean, and standard deviation.