Fig. 9. Oil saturation profiles for different stages of
flooding at the end of 0, 30, 60, 208, 294, 444 min in case of:(a) 14-6-14 GS; (b) 14-6-14 GS + PHPA; and(c) 14-6-14 GS + PHPA + SiO2.
Relative permeability curve analysis
In the presence of multiple fluid systems such as oil and water,
relative permeability describes the alteration in flow behavior with
saturation change. This is commonly observed during secondary and
tertiary flooding studies involving chemical induced displacement of
oil/water in porous media applications. Wettability alteration, pore
morphology, fluid distribution and saturation data are primary
parameters that influence relative permeability measurements
[66,67]. Figs. 10(a), 10(b) and 10(c) show the water/oil relative
permeability plots obtained during different flooding simulations. Solid
lines represent experimental data curves, wherein adjusted curves
obtained after CMOST assisted history matching are represented with help
of dotted lines. Initially, the core reservoir is in intermediate-wet
state, in which rock pore surfaces are wetted with oil and water exists
within the central regions between the pores. It is evident that
relative permeabilities for oil and water phases vary significantly with
increasing water saturation, which is brought about during
brine/chemical injection [67]. In summary, simulation studies prove
that the wetting nature of rock is altered to strongly water-wet state.
Fig. 10. Relative permeability curves of different case
scenarios involving: (a) 14-6-14 GS; (b) 14-6-14 GS +
PHPA; and (c) 14-6-14 GS + PHPA + SiO2flooding.
Cumulative oil production during secondary and tertiary
flooding
The oil recovery performance of surfactant, surfactant-polymer and
surfactant-polymer-nanoparticle based aqueous chemical fluids were
corroborated by history matching of experimental data. In previous
section, the laboratory results of flooding studies were discussed for
different chemical formulations (see Fig. 6). However, it is important
to study the validity of these results with compositional fluid flow
simulations such as CMG [17,18]. Fig. 11 shows a good match between
experimental and simulated outcomes of different flooding scenarios.
Experimental analyses revealed that water-flooding processes extracted
respective volumes of 5.73 cm3, 5.74
cm3 and 5.93 cm3 of initial oil
content in case scenarios I, II and II respectively. Simulation studies
showed water-flood recoveries of 44.86%, 44.31% and 44.99%
respectively at the end of secondary recovery, which is in close
agreement with experimental results. Simulation studies on tertiary
flooding studies showed crude oil recoveries of 7.65
cm3, 8.01 cm3 and 8.41
cm3 in the presence of {14-6-14 GS + chase water},
{14-6-14 GS + PHPA + chase water} and {14-6-14 GS + PHPA +
SiO2 chase water} respectively.
Fig. 11. Cumulative oil production versus time plots for:(a) 14-6-14 GS; (b) 14-6-14 GS + PHPA; (c)14-6-14 GS + PHPA + SiO2 showing match between
experimental and CMG-STARS results.
The error between experimental and simulated results for cumulative
flooding studies was obtained with the help of CMG-DECE (Designed
Exploration Controlled. Evolution) engine with 2000 experiments. Figs.
12(a), 12(b) and 12(c) presents the global history match (HM) error
versus experiment ID plots for cases I, II and III respectively. It is
evident that the simulation models were tailored during history match to
achieve optimized result(s) in the search direction of minimal error.
Error percentages with values ≤ 6.00% between the history matched and
experimental models was achieved during the simulation run. Cases I, II
and III registered the most optimal results for experiment ID nos. 1791
(within ± 5.85% error), 1017 (± 4.38 % error) and 1753 (± 5.23%
error) respectively. This optimized model was validated to match imput
fluid and rock conditions, and used as the best-fitted model to explain
flooding performances of different fluid systems. Table 5 shows the
rock-fluid parameters and flooding results for different case scenarios.
The parametric results were obtained from careful analysis of histogram
plots over significant number of simulation runs. The recovery rates,
oil/water permeability curves, rock-wetting properties and fluid flow
parameters were tuned during the history matching process. The study
ultimately presents the success of surfactant flood, surfactant-polymer
flood and surfactant-polymer-nanoparticle flood models in simulating the
experimental outcomes, and confirms their relative efficiencies
[17,18,26,27,35].
Fig. 12. History matching error between experimental and
simulated models, showing the base case, general solutions and optimal
solution for cumulative flooding characterized by : (a) 14-6-14
GS; (b) 14-6-14 GS + PHPA; (c) 14-6-14 GS + PHPA +
SiO2