Experimental Limestone Dissolution and Changes in Multiscale Structure Using Small- and Ultrasmall-Angle Neutron Scattering

Small-angle
neutron scattering (SANS), ultrasmall-angle neutron
scattering (USANS), backscatter electron (BSE) imaging, and neutron
computed tomography (NCT) were applied to the study of the pore size,
pore distribution, and pore connectivity developed during the experimental
dissolution of limestone. Eight cores of Indiana limestone having
initial permeabilities of 2–4 and 70 mD were reacted with HCl
solutions having a pH of 2 or 4 at flow rates of 0.1 or 10 cm3/min. NCT was used to image the structures developed during
dissolution. Nine cross sections of each core from the inlet to the
outlet were analyzed with SANS and USANS and with BSE imaging to characterize
changes in the pore structure throughout the length of the core after
reaction. The scattering curves obtained from SANS and USANS were
combined with autocorrelation analysis of the BSE images to characterize
porosity over length scales from ∼5 mm to 1 nm. Surface-fractal
dimensions were ∼2.3 in the nanopore region, and mass-fractal
dimensions were ∼2.75 in the micropore region. The transition
from surface- to mass-dominated fractal geometry is at a pore size
of ∼100 nm. There was no change in fractal behavior with dissolution,
pH, permeability, or flow rate. Porosity was generally greater at
the inlet, where most of the dissolution occurred, than at the outlet,
where there was little or no reaction. There was also some evidence
for porosity reduction near the inlet. The distribution of pore sizes
peaked in terms of pore numbers in the nano, micro, and meso range,
but there was little change in that distribution with dissolution.
There was also little change in porosity in samples that developed
preferential flow paths (wormholes), which formed in solutions of
low pH and, in particular, at high flow rate. The initial permeability
of each sample controlled the penetration and degree of branching
of each wormhole into the cores. Samples with wormholes had little
additional reaction. The composition of the solutions having a starting
pH of 4.0 approached the equilibrium value of 9.5 at the outlet, with
little regard to flow rate or permeability. In our experiments, the
formation of wormholes and the change in porosity were most strongly
influenced by the pH of the infiltrating solution, followed by the
flow rate and initial permeability.