Fluids and Rocks - Introduction to Carbon Capture and Sequestration - page 431

To understand why S g,r values are so challenging to characterize, let
us consider the capillary pressure curve of a generic rock sample. If we
knew the pore radius R and the mineral-water-CO 2 wetting angle

of
each individual pore, we could predict the capillary pressure curve of our
sample in the ideal case where each pore is fi lled with brine if P c < (2/ R )
γ gw cos

θ

and fi lled with CO 2 otherwise. This “ideal” P c curve (shown in
Figure 9.7.6 ) assumes that the fl uid fi lling each pore depends only on the
size of the pore and the wettability of the pore walls and does not depend
on the larger-scale structure of the pore network.
In reality, as noted in Figure 9.7.2 , measured P c curves have a sig-
nifi cant hysteresis, i.e., the P c curve is different during CO 2 invasion than
during brine imbibition. One reason for this hysteresis is that, whereas
brine imbibition in a pore is controlled by the pore size, CO 2 invasion in
a pore is controlled by the pore throat aperture. Another reason is that
the fi nite connectivity of pore networks gives rise to situations where, for
example, a large pore may be surrounded by smaller pores: during CO 2
invasion, this large pore will fi ll with CO 2 at a P c value that is determined,
not by its own size, but by the size of the surrounding smaller pores. As
a fi rst approximation, the net outcome of these effects is that the ideal P c
curve in Figure 9.7.6 is shifted toward higher P c values during CO 2 inva-
sion and toward lower P c values during brine imbibition.
As can be seen in Figure 9.7.6 , shifting the ideal P c curve upward
does not fundamentally change the overall character of the capillary inva-
sion process, whereas shifting the ideal curve downward allows the
emergence of a new phenomenon: residual CO 2 trapping. According to
Figure 9.7.6 , one might expect that the conditions that favor residual
trapping include an almost fl at P c curve (i.e., the porous medium must
contain many large pores) and a large hysteresis (i.e., the large pores
must be surrounded by narrow pore throats or smaller pores). The degree
by which the ideal P c curve is shifted downward during brine imbibition
(which has a very large impact on S g, r ) cannot be predicted solely from
the properties of individual pores: it requires an understanding of the
relative size of the pores and pore throats, the structure of the pore net-
work, and the fl uid dynamics of brine imbibition in the porous medium.
This fundamental difference between CO 2 invasion and brine imbibi-
tion is illustrated in Figures 9.7.7 and 9.7.8 . The fi rst fi gure shows (on the
right side) an X-ray micro-CT (computed tomography) image of CO 2
distribution during CO 2 invasion in a sandstone sample. The fi nely

θ