Large-Scale Geological Carbon Sequestration - Introduction to Carbon Capture and Sequestration - page 495

Air Air

Gas

Gas

(a) (b)

Air Air

(c) (d)

Figure 10.4.7 Density-dependence in mixing of two gasses
Thought experiment elucidating the role of density-dependence in mixing of two gasses.
We start with a container of gas on a fl at surface at ambient conditions (a). We then
remove the walls of the container (b). With no wind, a passive gas will mix only by diffu-
sion (c) whereas a dense gas will fl ow on its own and mix by dispersion (d).

dispersion by wind is quite effective, which leads to the conclusion that
above-ground HSE impacts due to leaking CO 2 are primarily local and
will not extend very far downstream due to turbulent mixing. This result
is corroborated by more recent simulation studies of pipeline leakage
involving very accurate pipeline leakage fl ux evolution and three-dimen-
sional Navier-Stokes atmospheric dispersion processes [10.46].
From these studies, it appears that density effects in a CO 2 surface
leakage scenario only need to be considered if the seepage area is large
or the wind is very light. This fi nding is quantifi ed in Figure 10.4.8 through
application of the Britter and McQuaid [10.45] density-dependence crite-
rion for a variety of system properties. As shown in the fi gure, seepage
fl uxes of the order 10 4 times larger than those of the natural ecosystem
exchange will be density-dependent, even for relatively large wind
speeds, while lesser fl uxes are often passive by this criterion.
In summary, there are recognized hazards of CO 2 injection, but they
are generally local effects and they are considered manageable. The
hazards of geological carbon sequestration should always be considered
relative to other energy-related hazards (see Question 10.4.1 ) and most