Mass transfer between a multicomponent trapped gas phase and a mobile water phase: Experiment and theory

Gas tracer experiments were carried out in dynamically compressed sediments to investigate the mass transfer between a trapped multicomponent gas phase and a mobile water phase. The saturation state of the column was characterized by three independent methods: (1) by gravimetric measurements, (2) by bromide tracer tests, and (3) by hydraulic conductivity measurements. For inverse modeling a new kinetic model was developed allowing volume change of the entrapped gas. The new kinetic model consistently explains oxygen elution curves, the time evolution of the integral gas saturation, and integral hydraulic conductivity. The sensitivity of three different velocity-dependent mass transfer correlations to the dissolution process was investigated: (1) a classical square-root, single-sphere correlation, Sh similar to Pe(0.5), (2) a multisphere correlation, Sh similar to Pe(n) (n = 0.5-1.0), and (3) an empirical correlation, Sh similar to Pe(0.8). It was found that all correlations yield nearly the same elution curves for 10 gas tracer experiments with three different two-component gas phases: O-2/He, O-2/N-2, and O-2/Ar and for different flow velocities ranging from 5 to 20 m d(-1). For all gas tracer experiments a distinct minimum of the longitudinal dispersivity was found during gas dissolution, i. e., in the unsaturated state. For the saturated state we found that the experimental values could be described by Saffman's theory: (D) over tilde (p) proportional to Pe ln (Pe) with a normalized mean square root error of 6%.