Nitin Chowdhury's thesis
by
Nitin Chowdhury, MSE
University of Texas at Austin, 2003
Supervisors: Gary A. Pope
Kamy Sepehrnoori
The main objective of this work was to use compositional simulation
to better predict the well and reservoir performance for gas condensate
reservoirs and reservoirs undergoing asphaltene precipitation. The first part
of this work deals with simulation of asphaltene precipitation in reservoirs
and the prediction of damage caused by it. A new semi-analytical method was
developed in the second part of this work to predict the well deliverability in
gas condensate reservoirs with much less computational cost.
Asphaltene precipitation in reservoirs can cause plugging of the
pore space and thus, reduce the permeability of the formation. The models used
for the simulation of asphaltene precipitation in the GEM and UTCOMP simulators
were reviewed and a comparison was done between the two simulators. The damage
caused due to asphaltene precipitation in reservoirs was studied in terms of
the permeability reduction and the effect that it has on further asphaltene
precipitation. This work was focused on studying asphaltene precipitation with
respect to pressure. The calibration of
the asphaltene precipitation model and the permeability reduction model to
experimental data is very important before it can be used for making any useful
predictions. A sensitivity study was performed to investigate the effect of
varying the permeability reduction and adsorption of the asphaltene on the
rock. The adsorption of the asphaltene on the rock is found to be an important
parameter controlling the damage caused by asphaltene precipitation. A layered
reservoir with a high-permeability contrast was simulated to investigate the
effect of heterogeneity on asphaltene precipitation. Asphaltene precipitation
first occurred in the high permeability layers in this simulation. This reduced
the preferential flow through the high-permeability layers and increased the swept volume.
The second part of the thesis is focused on accurately predicting
the deliverability for gas condensate wells without the need of performing
computationally expensive fine-grid simulations. A fine-grid simulation is
generally needed to capture the build up of the condensate bank near the well
bore and also any change in the physical properties in the condensate bank.
However, fine-grid simulations suffer from the disadvantage of large run times
that makes their use unrealistic for full-field simulations. A new
semi-analytical method was developed that can be used in conjunction with a
coarse-grid simulation to accurately predict the gas and oil production rates
for wells in a gas-condensate field. The theory and the underlying assumptions
behind the method are discussed in detail.High velocities and low interfacial tension
between gas and condensate causes the gas relative permeability to increase,
which leads to higher production rates than would otherwise be true. This
effect was modeled in this work with a trapping number model. The
non-Darcy flow may also be significant at high flow rates. The new method
captures both of these near-well effects that are critical to accurate
calculation of the well productivity.
The new method was coded into the compositional reservoir
simulator UTCOMP and compared against fine-grid compositional simulation
results for both lean and rich gas condensate fluids. A detailed comparison of
near-well saturations, relative permeabilities, viscosities and densities with
fine-grid simulation results is presented for verification of the new method.
The method was tested for single-layer, multi-layer and multi-well
gas-condensate reservoirs and found to give accurate results compared to
fine-grid simulations.
Back to theses index
