On the response of tracer tests to fracture aperture variability and its correlation with fracture surface area

Fractures are relevant for many subsurface applications such as geothermal energy, radioactive waste disposal or petroleum engineering. The exchange of heat or chemical species between the rock and the fluid is an important process for many of these applications. This exchange strongly depends on the effective interfacial area between the flowing fluid and the rock, i.e. the effective fracture surface area. In the field, it is common practice to perform tracer tests to obtain reservoir parameters such as the fracture surface area. However, the fracture surface area is also one of the most uncertain parameters in the characterization of fractured reservoirs.

Generally, fractures of all scales can be characterized by two opposite rock surfaces with variable surface roughness. Consequences of varying surface roughness are a heterogeneous aperture distribution and the formation of channels and barriers within the fracture planes. This significantly affects flow and transport and the effective fracture surface area. Hence, the aperture distribution is highly relevant for exchange processes.

In this study, we represent local hydraulic aperture statistically by a spatially autocorrelated random field, which is characterized by the variogram model and a correlation length. The focus of the study is to investigate the influence of this correlation length on the effective fracture surface area and its implication for the results of tracer tests. To this end, numerical simulations of a single circular fracture are performed. In the fracture plane, flow and tracer transport occur between an injection and a production well.

We calculate the swept fracture volume based on the tracer mean residence time and the interwell volumetric flow rate. Given this swept volume, the effective fracture surface area can be determined from the simulations, revealing the processes responsible for the characteristics of the tracer breakthrough curves. Furthermore, our simulations enable the investigation of the influence of different correlation lengths on the results of tracer tests. We show how the aperture variability impacts the results of tracer tests and the fractured surface area. Our results contribute to the improvement of fractured reservoir characterization.