Integrating Tracer Test Data into Geostatistical Aquifer Characterization at the Hanford 300 Area

Tracer testing under natural or forced gradient flow is an efficient method for characterizing subsurface properties, by monitoring and modeling the tracer plume migration in a heterogeneous aquifer. At the Hanford 300 area, non-reactive tracer experiments, in addition to constant-rate injection tests and electromagnetic borehole flow meter profiling, were conducted to characterize the heterogeneous hydraulic conductivity field at the site. The tracer was injected at a near-constant rate for 10 hours, and the tracer concentrations were monitored for 12 days in a network of observation wells. This work presents a Bayesian inverse modeling technique to infer the heterogeneity structure of the hydraulic conductivity in the saturated zone of the Hanford formation, using the breakthrough curves at various observation wells. Analytical or semi-analytical solutions for mass transport in divergent radial flow fields are adopted whenever possible to avoid expensive numerical forward simulations. Compared to the case conditioned on the constant-rate injection tests and electromagnetic borehole flow meter profiling, this study finds that the inclusion of tracer test data can improve the estimation of heterogeneity structure and reduce the prediction uncertainty of the solute transport at given locations. With the availability of observation wells at varying distances, we also investigate the worth of data in each observation well, which can be used to evaluate the effectiveness of current experimental setup and guide further data collection practices at the site.