|Abstract: ||Understanding the displacement and trapping of a displaced phase in porous media is
important for applications in improved oil recovery (IOR) and carbon capture and storage
(CCS). In IOR, we design the process to leave as little residual oil behind as possible,
while for CCS, we do the opposite: we wish to maximise the amount of CO2 trapped by
the host brine. Reservoir rocks display a range of wettability, from being preferentially
water-wet–they spontaneously imbibe water–to oil-wet, or water repellent. Many rocks
are mixed-wet, with both water-wet and oil-wet pores. The other wettability state is
more intermediate-wet where, as we show, the rock appears to be largely non-wetting
to both oil and water. Carbonate reservoirs, which house the majority of the world’s
remaining conventional oil, and which offer potential storage locations for carbon dioxide,
have an altered wettability after contact with crude oil. In this thesis we study
spontaneous displacement and trapping in carbonate rocks for different wettability conditions.
The rate of spontaneous imbibition governs the rate with which oil, or carbon
dioxide is trapped, while the residual saturation quantifies how much trapped. This is
particularly important in carbonate reservoirs, which are almost extensively fractured.
In these reservoirs, the principal mechanism for displacement is spontaneous imbibition
of water to displace oil (or carbon dioxide) in the water-wet portions of the pore space.
Pore structure and wettability are two of the main factors affecting displacement and
capillary trapping. Experimental and pore-scale modelling studies have found a monotonic
increase of residual non-wetting phase saturation, Snwr, with the initial non-wetting
phase saturation, Snwi in a water-wet medium. However, altered-wettability systems
have received relatively little attention, particularly those which are intermediate-wet.
We first present the three carbonates we study in this thesis: Estaillades, Ketton and
Portland. These are three quarry limestones that have very different pore structures and
span a wide range of permeability. We present standard core analysis results including
mercury injection capillary pressure and nuclear magnetic resonance response. We also
study three-dimensional X-ray images of these samples, obtained at a resolution of a few
microns. We use these experiments to assess the pore size distribution; we show that
all the samples have micro-porosity and use the results to interpret the trapping and
displacement experiments performed later.
We then perform spontaneous imbibition experiments in these three carbonates under
strongly water-wet conditions. We use scaling equations and recently published analytical solutions to assess the recovery of these rocks. We perform two sets of experiments.
In the first, we measure the mass of water imbibed as a function of time. We show that
the amount imbibed scales as the square root of time. In the second series of experiments,
we measure saturation profiles as a function of distance and time using X-ray CT
scanning. We demonstrate that the saturation profiles are functions of distance divided
by the square root of time. We also demonstrate that the profiles are consistent with
the analytical theory and, using reasonable estimates of relative permeability and capillary
pressure, we can match the experimental results with the analytical solutions. We
discuss how, in combination with conventional measurements of relative permeability
(steady-state or using Buckley-Leverett theory in an unsteady-state experiment) these
measurements could be used to measure capillary pressure and relative permeability.
In the next phase of the study, we use organic acid (cyclohexanepentanoic acid) to
alter the wettability of our samples and observe the relationship between the initial oil
saturation and the residual saturation. We take cores containing oil and a specified
initial water saturation and waterflood until 10 pore volumes have been injected. We
record the remaining oil saturation as a function of the amount of water injected. In
the water-wet case, with no wettability alteration, we observe, as expected, a monotonic
increase between the initial and the remaining oil saturation. However, when the wettability
is altered, we observe an increase, then a decrease, and finally an increase in the
trapping curve for Estaillades limestone with a small, but continued, decrease in the remaining
saturation as more water is injected. This behaviour is indicative of mixed-wet
or intermediate-wet conditions as there is no spontaneous imbibition of oil and water.
However, Ketton did not show indications of a significant wettability alteration with
a similar observed trapping profile to that observed in the water-wet case. Portland
limestone also showed a monotonic increasing trend in residual saturation with initial
saturation but with a higher recovery, less trapping, than the water-wet case. Again,
this is intermediate-wet behaviour with no spontaneous imbibition of either oil or water,
and slow production of oil after water breakthrough. Finally, we repeat the same experiments
but instead we age the three carbonates with a high asphaltenic content and high
viscosity crude oil at 70C mimicking reservoir conditions. The results show a monotonic
increase in residual saturation as a function of initial saturation but with higher recovery
than the water-wet cases for Estaillades and Portland, with again no indication of
wettability alteration for Ketton. We discuss the results in terms of pore-scale recovery
process and contact angle hysteresis. We observe recovery behaviour that lies between
the water-wet and mixed-wet conditions previously studied in the literature.
Overall, the thesis demonstrates that recovery rate and the amount of trapping are
sensitive to pore structure and wettability. Very different recovery trends were observed
for three rocks with similar chemical composition. The work serves as a benchmark for
further modelling and experimental studies. The recommendation is to reproduce, in
the laboratory, conditions close to those observed in the reservoir, and to use imbibition
and displacement measurements to quantify and constrain multiphase flow properties.|