Tracer Transport in Fractures: Analysis of Field Data Based on a Variable‐Aperture Channel Model

Abstract

A variable‐aperture channel model is used as the basis to interpret data from a 3‐year tracer transport experiment in fractured rocks. The data come from the so‐called Stripa‐3D experiment performed by Neretnieks and coworkers. Within the framework of the variable‐aperture channel conceptual model, tracers are envisioned as traveling along a number of variable‐aperture flow channels, whose properties are related to the mean and standard deviation σ of the fracture aperture distribution. Two methods are developed to address the presence of strong time variation of the tracer injection flow rate in this experiment. The first approximates the early part of the injection history by an exponential decay function and is applicable to the early time tracer breakthrough data. The second is a deconvolution method involving the use of Toeplitz matrices and is applicable over the complete period of variable injection of the tracers. Both methods give consistent results. These results include not only estimates of and σ, but also the number of channels involved in the tracer transport and their Peclet numbers and dispersivities. An interesting and surprising observation is that the data indicate that for each channel the Peclet number increases with the mean travel time; i.e., dispersivity decreases with mean travel time. The meaning of this trend is discussed in terms of the strong heterogeneity of the flow system.