A refined approach to estimating effective flow porosity from cross-hole tracer tests in fractured media

Simulations of flow and transport in two-dimensional representations of heterogeneous fractured media are used to investigate the errors and biases associated with effective flow porosity estimates derived from cross-hole tracer tests. A method is presented for constructing probability distributions of 'correction factors' that can be used to correct apparent flow porosities obtained from tracer tests to obtain 'true' flow porosities in fracture systems. Although only a limited number of the many possible variations in fracture flow system properties is investigated, it is concluded that effective flow porosities derived from cross-hole tracer tests have a strong tendency to overpredict true flow porosities in fracture flow systems. This tendency toward overprediction decreases as the fracture conductivity relative to the background conductivity field decreases and as the orientation of the most conductive fractures becomes better aligned with the two wells. Tracer tests with small amounts of recirculation of water from the production well to the injection well are predicted to result in much better estimates of true flow porosity (on average), and with much less variability in the estimates, than tests with no reclrculation. However, the advantage offered by recirculation decreases as the fracture conductivity relative to the background conductivity decreases.