Intrinsic Dispersivity of Randomly Packed Monodisperse Spheres

Abstract

We determine the intrinsic longitudinal dispersivity l(d) of randomly packed monodisperse spheres by separating the intrinsic stochastic dispersivity l(d) from dispersion by unavoidable sample dependent flow heterogeneities. The measured l(d), scaled by the hydrodynamic radius r(h), coincide with theoretical predictions [Saffman, J. Fluid Mech. 7, 194 (1960)] for dispersion in an isotropic random network of identical capillaries of length l and radius a, for l/a=3.82, and with rescaled simulation results [Maier et al., Phys. Fluids 12, 2065 (2000)].

Enhanced gas recovery with CO2 sequestration: The effect of medium heterogeneity on the dispersion of supercritical CO2–CH4

“…In general, carbonate rocks are characterized as a more heterogeneous media than sandstone rocks and therefore require a longer time to reach the asymptotic regime. Recently, there have been extensive studies on pre-asymptotic dispersion in various porous media (Bijeljic et al, 2011(Bijeljic et al, , 2013aBijeljic et al, 2013b;Blunt et al, 2013;Mitchell et al, 2008;Scheven et al, 2007;Scheven et al, 2005). Most of these focused on velocities in the transition region or above is Fickian-like and predominately dictated by restricted diffusion (Bijeljic et al, 2004;Gist et al, 1990;Levy and Berkowitz, 2003;Sahimi, 1995).…”

Magnetic resonance imaging measurements evidence weak dispersion in homogeneous porous media

“…Such difference may be attributed to quite limited explored length scales, so that some macroscale effects responsible for higher macroscale dispersivity may be missed. Besides, from MRI (Magnetic Resonance Imaging) of D 2 O injection in chromatography gel columns a significant band broadening due to column inlet was observed [12], and from an analysis of NMR data, it was suggested that imperfect flow injection could also have a significant impact on the dispersion observed from breakthrough curve experiments [22]. This work also showed that a further analysis of PFG data makes it possible to distinguish intrinsic dispersion, which appears to be in agreement with the theoretical prediction for a random network of capillaries [23], whose resulting D is below the master curve (see above) by a factor about 3.…”

An empirical correlation of the longitudinal and transverse dispersion coefficients for flow through random particle packs

“…However, a significant scatter exists. The scatter may be caused by the difference of the pore geometry and the flow heterogeneities during the solute injection process in different experiments (Scheven et al, 2007). The packing method, wall confinement, particle size and shape distributions all may influence the pore geometry.…”

“…The scatter hinders our accurate understanding of the dispersion behavior. Recently, an important advancement has been made (Scheven et al, 2007(Scheven et al, , 2014Scheven, 2013). The longitudinal (Scheven et al, 2007(Scheven et al, , 2014 and transverse (Scheven, 2013;Scheven et al, 2014) dispersion coefficients for random particle packs were obtained by nuclear magnetic resonance (NMR) based experiments and the dispersion coefficients for columns with different pore geometries collapsed into the same value.…”

“…The longitudinal (Scheven et al, 2007(Scheven et al, , 2014 and transverse (Scheven, 2013;Scheven et al, 2014) dispersion coefficients for random particle packs were obtained by nuclear magnetic resonance (NMR) based experiments and the dispersion coefficients for columns with different pore geometries collapsed into the same value. The dispersion caused by the unavoidable macroscopic flow heterogeneities was separated from the intrinsic dispersion by their data analysis (Scheven et al, 2007(Scheven et al, , 2014Scheven, 2013). Numerical simulation also reproduced the experimental results (Scheven et al, 2014).…”