High Pressure Air Injection (HPAI) is an improved oil recovery process in which
compressed air is injected into typically deep, light oil reservoirs. Part of
the oil reacts exothermically with the oxygen in the air to produce flue gas
(mainly composed of nitrogen, carbon dioxide and water). Literature explaining
the reaction mechanisms and phase interactions is available. Nevertheless,
little effort has been devoted to describing gas, oil and water three-phase
flow behaviour under HPAI reservoirconditions.
Three coreflood experiments were conducted on Berea sandstone core. The first
experiment consisted of injecting flue gas into core at initial oil and connate
water saturations to obtain liquid-gas relative permeability data. The second
experiment was designed to evaluate oil re-saturation, after gas sweep,
simulating an HPAI thermal front. The third experiment consisted of gas
displacing both oil and water completing the data necessary to plot the
three-phase relative permeability curves.
Reservoir simulation was used to adjust relative permeability curves and
hysteresis parameters by matching the pressure drop and production data.
It is well-known that the oil recovery mechanisms in HPAI are a combination of
highly efficient displacement by the reaction front and light oil/flue gas
compositional interactions, such as oil swelling and/or vapourization and
near-miscible behaviour(1). However, the contribution of each of
these recovery mechanisms has not been properly assessed(2).
Although significant effort has been devoted to the characterization of
oxidation kinetics(3-5) and flue gas/light oil compositional
interactions(6, 7), the process remains challenging to simulate even
under controlled and ideal conditions, i.e., a combustion tube test. Some of
the difficulties include the limited availability of experimental data to feed
the numerical simulators with the required parameters, as well as the
interdependence of these parameters and their variation with temperature.
Assuming that these difficulties can be overcome by carrying out a study that
allows a judicious analysis of experimental information and a careful treatment
of the matched parameters in a numerical simulator, there is still a piece of
information that has a strong influence on the simulation results and cannot be
defaulted or left as a final matching tool: relative permeability.
In an earlier study(2), it was suggested that for a combustion tube
match, the steps previous to air injection (waterflood and inert gas flood)
could be used to find a full set of relative permeability curves for the run.
It was also pointed out that the rock-fluid dataset should include a variation
of relative permeability data with interfacial tension to account for changes
in pressure, composition, and most importantly, temperature. While this being
necessary, it is still insufficient to ensure a correct representation of the
flow of phases in a porous medium subjected to HPAI.
Ahead of the reaction front, the high mobility flue gas (mainly composed of
nitrogen and carbon oxides) displaces oil and water at nearly reservoir
temperature, while in the high temperature zone, oil and water are evapourated
to later condense downstream. In both zones, three-phase flow is occurring.
© 2009. Petroleum Society of Canada (now Society of Petroleum Engineers)
View full textPDF
- Original manuscript received:
27 March 2008
- Meeting paper published:
17 June 2008
- Revised manuscript received:
15 May 2009
- Manuscript approved:
4 August 2009