Summary
Interest in the vapor-extraction (Vapex) process for heavy-oil and bitumen
recovery has grown considerably as a viable and environmentally friendly
alternative to the currently used thermal methods. The potential for the
success of the Vapex process is even more attractive in some scenarios that
preclude the thermal methods. The presence of an overlying gas cap and/or
bottomwater aquifer, thin pay zones, low thermal conductivity, high water
saturation, and unacceptable heat losses to overburden and underburden
formations are some of the limitations with the thermal techniques, which
potentially can be overcome by Vapex implementation. However, predicted low
production rates by previous researchers for field application of the Vapex
technique remain a serious barrier to commercial applications of the process.
The scaleup methods that have been used by previous workers for translating the
laboratory results to field predictions were based primarily on the reservoir
transmissibility. An analytical model developed by Butler and Mokrys showed
that the oil rate should be proportional to the square root of reservoir
transmissibility. The effect of convective dispersion between solvent and
virgin heavy oil in porous media was ignored in developing this
model.
The main objective of this work is to develop an improved scaleup method for
the Vapex process using physical-model experiments carried out in models of
different sizes. In this paper, we report the results of a new series of
experiments that extend the previously reported results of Karmaker and
Maini to a significantly wider range of model heights. These new experiments
used a new design of slice-type physical models that places the sandpack in the
annulus formed by two cylindrical pipes. Combining the new results with the
previous data of Karmaker and Maini, we show that the transmissibility-based
scaling-up method seriously underpredicts the results at larger scales. This
observation suggests that much higher rates can be expected in the field
implementation of the Vapex process.
A new correlation also has been proposed for scaling up the experimental
data to the real field cases. It indicates the height dependency of the
convective-dispersion contribution, which can be the dominant mass-transfer
mechanism for the process, to be a higher order than previously postulated.
Experimental results from this work show that the stabilized rate is a function
of drainage height to the power of 1.1 to 1.3, instead of the square-root
functionality of the Butler and Mokrys model.
Introduction
Cost-effective heavy-oil- and bitumen-recovery methods are still challenging
issues that have not been fully resolved. The huge volume of almost immobile
hydrocarbon resources in the world, especially located in Canada, Venezuela,
and the United States (approximately six times the total conventional oil
reserves), offers unlimited challenges and opportunities to researchers. The
high viscosity and low mobility of these oils cause the primary recovery to be
very low. The adverse mobility-ratio problem also limits the application of
waterflooding to these reservoirs. The overall recovery that can be achieved
before the enhanced-oil-recovery (EOR) methods usually does not exceed 6 to 8%
of the original oil in place.
The well-known observation of a dramatic decrease in the viscosity of heavy
oil with temperature increase makes the thermal-recovery methods, such as
steamflooding, cyclic steam stimulation (CSS), in-situ combustion, and (more
recently) the steam-assisted gravity drainage (SAGD) process the obvious
choices. However, thermal methods are not universally applicable to highly
viscous heavy-oil reservoirs. The low recovery factors associated with CSS
(inefficient steamflood in highly viscous oils and a relatively high mobility
requirement), in addition to the process-control difficulties for the in-situ
combustion technique, are some of the obstacles that leave the SAGD process as
the only thermal option for heavy-oil and bitumen recovery in many reservoirs.
In the SAGD process, two horizontal wells located in the same vertical plane
are used to inject the steam from the upper well and produce heated oil from
the lower well.
The Vapex process, which was initially proposed by Butler and Mokrys, is a
solvent-based analog of the SAGD process, which can be considered when the SAGD
is likely to be problematic. In thin reservoirs, the amount of heat loss to the
surrounding formations makes the SAGD uneconomic. Also, in low-permeability
carbonate reservoirs in which the heat capacity per volume of oil is high, the
steam/oil ratio is not economically attractive. The presence of the bottom
aquifer and/or a thin gas cap can be counted as an advantage for the Vapex
process, whereas they are troublesome for SAGD. In terms of energy
consideration, it has been reported that Vapex needs only a fraction of the
energy used for SAGD. Also, Vapex has smaller upfront capital requirements
compared to SAGD, in which 30% of the capital investment goes toward
steam-generation equipment.
© 2005. Society of Petroleum Engineers
View full textPDF
(
2,344 KB
)
History
- Original manuscript received:
12 January 2004
- Revised manuscript received:
15 March 2005
- Manuscript approved:
31 March 2005
- Version of record:
15 June 2005
View Cart
