Insight from simulations of single-well injection-withdrawal tracer tests on simple and complex fractures

The single-well injection withdrawal (SWIW) test, a tracer test utilizing only one well, is proposed as a useful contribution to site characterization of fractured rock, as well as providing parameters relevant to tracer diffusion and sorption. The usual conceptual model of flow and solute transport through fractured rock with low matrix permeability involves solute advection and dispersion through a fracture network coupled with diffusion and sorption into the surrounding rock matrix. Unlike two-well tracer tests, results of SWIW tests are ideally independent of advective heterogeneity, channeling and flow dimension, and, instead, focus on diffusive and sorptive characteristics of tracer (solute) transport. Thus, they can be used specifically to study such characteristics and evaluate the diffusive parameters associated with tracer transport through fractured media. We conduct simulations of SWIW tests on simple and complex fracture models, the latter being defined as having two subfractures with altered rock blocks in between and gouge material in their apertures. Using parameters from the Aspo site in Sweden, we calculate and study SWIW tracer breakthrough curves (BTCs) from a test involving four days of injection and then withdrawal. By examining the peak concentration C{sub pk} of the SWIW BTCs for a variety of parameters, we confirm that C{sub pk} is largely insensitive to the fracture advective flow properties, in particular to permeability heterogeneity over the fracture plane or to subdividing the flow into two subfractures in the third dimension orthogonal to the fracture plane. The peak arrival time t{sub pk} is not a function of fracture or rock properties, but is controlled by the time schedule of the SWIW test. The study shows that the SWIW test is useful for the study of tracer diffusion-sorption processes, including the effect of the so-called flow-wetted surface (FWS) of the fracture. Calculations with schematic models with different FWS values are conducted and the possibility of direct in situ measurement of FWS with SWIW tests is demonstrated