Effects from influent boundary conditions on tracer migration and spatial variability features in intermediate-scale experiments

In previous unsaturated transport studies at Los Alamos dispersion coefficients were estimated to be higher close to the tracer source than at greater distances from the source. Injection of tracers through discrete influent outlets could have accounted for those higher dispersions. Also, a lack of conservation of mass of the tracers was observed and suspected to be due to spatial variability in transport. In the present study experiments were performed under uniform influent (ponded) conditions in which breakthrough of tracers was monitored at four locations at each of four depths. All other conditions were similar to those of the unsaturated transport experiments. A comparison of results from these two sets of experiments indicates differences in the parameter estimates. Estimates were made for the dispersion coefficient and the retardation factor by the one-dimensional steady flow computer code, CFITIM. Estimates were also made for mass and for velocity and the dispersion coefficient by the method of moments. The dispersion coefficient decreased with depth under discrete influent application and increased with depth under ponded influent application. Retardation was predicted better under the discrete influent application than under ponded influent application. Differences in breakthroughs and in estimated parameters among locations at the same depth were observed under ponded influent application. Those differences indicate that there is a lack of conservation of mass as well as significant spatial variability across the experimental domain. 14 refs., 9 figs., 8 tabs