Flow dimension as an indicator of hydraulic behavior in site characterization of fractured rock

We examine the possibility of using the flow dimension identified from constant pressure injection tests as a tool for characterizing the hydraulic conditions of fractured media. The data comes from a low-conductivity crystalline rock site, from depths of up to 450 m, and is obtained with 2 m and 10 m measurement scales. In the analysis, the general solution for n-dimensional flow by Barker (1988) is applied. The results show that the most prominent characteristics of the medium can be identified; that is, linear and sublinear flow dimensions as distinguished from dimensions higher than two. In many cases, however, there is significant difficulty in distinguishing the dimensions n = 2, 2.5, and 3 from each other. This is usually because of the experimental difficulties in achieving the ideal conditions required by the theory during the early part of the experiment. In such cases, a full flow curve is not available for the type-curve fitting. In the nonunique cases the higher dimensions typically correspond to higher, sometimes unrealistically high, values of specific storage and to the less reliable and less representative early part of the experiment. Therefore, most of the dimensions in categories n = 3 can be excluded, thus leaving the majority observations in the categories of n = 2 and n = 2-2.5. The dominance of dimension n = 2 is more pronounced for data related to fracture zones in comparison to that related to "average" rock, in particular in the 2 m scale data. The proportion of low (n < 1.5) flow dimensions is small, but for the 10 m scale data it is relatively higher at greater depths and corresponds to lower conductivities. For the smaller 2 m scale data, the low dimensions are not linked to greater depths or systematically smaller conductivities, giving preliminary indication of different flow dimension behavior for the two different scales.