Summary
Equation-of-state (EOS) compositional and surfactant models are coupled in a
fully implicit parallel reservoir simulator using the equivalent alkane carbon
number (EACN) of the oleic phase. The EACN of the oleic phase is computed using
a mole-fraction-weighted carbon number for each component present in the oleic
phase. Important microemulsion properties such as optimum salinity and optimum
solubilization parameter as a function of the EACN of the oleic phase are
implemented on the basis of known correlations. Type II(−) surfactant phase
behavior is considered in this study. The simulator developed is validated
using our implicit-pressure/explicit-concentration (IMPEC) chemical-flooding
simulator. Case studies, including a large-scale simulation, emphasize that
surfactant floods should be modeled carefully, taking the EACN of crude oil
into consideration for more realistic and accurate oil-recovery
predictions.
Introduction
Surfactant flooding is one of the most effective methods to improve oil
recovery: Dissolving oil in the aqueous phase results in lowering the
interfacial tension between the oleic and aqueous phases. It is often
accompanied by polymer flooding to increase sweep efficiency. The
surfactant/polymer flooding has become a more competitive process under the
current circumstances of high oil prices, and oil companies are considering the
process more seriously to rejuvenate their mature fields (Chang et al. 2006;
Anderson et al. 2006).
Phase behavior of a brine/oil/surfactant formulation is one of the key
factors determining the enhanced oil recovery by surfactant flooding. Early
experimental studies (Salager et al. 1979a; Salager et al. 1979b; Barakat et
al. 1983; Baran et al. 1994) have shown that the surfactant-phase behavior is a
strong function of the hydrocarbon composition of the crude oil. Surfactant
properties such as optimum salinity and solubilization parameters were found as
a function of the EACN of the oil. The EACN is a mole-fraction-weighted carbon
number for each component present in the oleic phase. For example, the EACN is
9.0 if the oil consists of octane, nonane, and decane with mole fraction of
0.3:0.4:0.3, respectively.
When chemical flooding is applied to a reservoir with either free or
dissolved gas, the oil composition may change considerably because of the mass
transfer between the gaseous and oleic phases. This change results in spatial
variation of the EACN of oil and, therefore, affects the surfactant-phase
behavior and, ultimately, the oil recovery and process performance. However,
the effect of oil composition or oil EACN on the surfactant-phase behavior
often has been neglected when modeling such processes by assigning constant
optimum salinity and solubilization parameters for the specific reservoir crude
oil and surfactant.
Few chemical-flooding simulators other than the University of Texas Chemical
Compositional Simulator, UTCHEM, have been developed to consider the effect of
spatial variation of the EACN on the surfactant-phase behavior. The simulator
is a 3D, multiphase, multicomponent chemical-flooding simulator and has been
used extensively and validated with laboratory and field data (Delshad et al.
1996). However, it is IMPEC code and, in its current form, cannot run on
parallel computers. Therefore, if a large number of gridblocks are necessary
for either simulation of a large reservoir or refinement of a model for more
accurate chemical-flooding simulation, run time can be very long because of the
timestep restriction. Also, computational memory of a single processor can be
insufficient for the large problem size. To overcome this computational
limitation, a fully implicit, parallel, EOS compositional chemical flooding
simulator, called the General Purpose Adaptive Simulator (GPAS), has been in
development (Han et al. 2007).
Currently, GPAS can model only two phases under optimum Type II(−)
surfactant phase behavior, which is explained in the next section. Several
other modeling features relating to chemical flooding, such as cation exchange
and chemical reactions, are not yet available in the simulator. However, with
the fully implicit scheme and the capability of parallel computation, it has
shown the ability to perform field-scale, high-resolution
surfactant/polymer-flood simulations with more than one million gridblocks and
taking large timesteps (Han et al. 2007).
The subject of this paper is the implementation of EACN and its effect on
surfactant phase behavior in GPAS. Currently, this implementation considers
only Type II(−) surfactant phase behavior. However, this work is the first and
crucial step toward a more accurate model of chemical phase behavior in oil
reservoirs with dissolved or free gas present and the processes that involve
the coinjection of gas and surfactant such as foam or surfactant alternating
gas (SAG).
© 2009. Society of Petroleum Engineers
View full textPDF
(
1,456 KB
)
History
- Original manuscript received:
3 July 2006
- Meeting paper published:
24 September 2006
- Revised manuscript received:
18 November 2008
- Manuscript approved:
8 December 2008
- Published online:
1 June 2009
- Version of record:
1 June 2009
View Cart
