Numerical
modeling of asphaltene precipitation in petroleum reservoirs
is important in relation to possible precipitation around the wellbore
in the producing well. Production from some reservoirs results in
asphaltene precipitation in the wellbore region, leading to productivity
loss and need for cleanup. Fluid injection even when there is asphaltene
precipitation may not lead to injectivity loss. There are desirable
processes in which precipitation of asphaltenes can lead to “in
situ” upgrading of heavy oil recovery. Reservoir compositional
models that are currently in use rely on cubic equations of state
for asphaltene precipitation. The cubic equations, despite their relative
reliability in describing reservoir fluids’ phase behavior,
become unreliable in asphaltene-rich phase description. A number of
noncubic equations of state have been introduced to overcome the shortcomings
of cubic equations. The cubic-plus-association equation of state (CPA-EOS)
is perhaps the method of choice in representing asphaltenes in compositional
modeling. When the hydrocarbon fluids do not contain asphaltenes,
CPA-EOS reduces to the standard cubic equation. In this work, we implement
CPA-EOS in compositional modeling and introduce a simple technique
to speed up considerably the root finding of the CPA-EOS. Our efficient
algorithm reduces significantly the additional computational cost
from the incorporation of the CPA-EOS. We also derive the basic equations
for the total compressibility and total partial molar volume in our
implementation of the CPA-EOS compositional modeling. We present three
numerical examples for CO2 injection in 2D and 3D domains
saturated with Weyburn oil and show results of asphaltene-rich phase
saturation among other predictions. This work introduces a general
framework for widespread use of CPA-EOS in compositional modeling
in three-phase flows of gas, light liquid, and asphaltene-rich phases.