Society of Petroleum Engineers

SPE 25452

In-Situ Upgrading of Heavy Oils and Bitumen by Propane

Deasphalting: The Vapex Process
I.J. Mokrys and R.M. Butler, U. of Calgary
SPE Members

Copyright 1993. Society of Petroleum Engineers. Inc.

This paper was prepared for presentation at the Production Operations Symposium held in Oklahoma City. OK. U.S.A., March 21-23, 1993.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A. Telex, 163245 SPEUT.

ABSTRACT INTRODUCTION

Many of the Lloydminster-type heavy oil reservoirs are A third of the world's oil is in Canada as heavy oil and bitumen.
unsuitable for thermal recovery by steam injection because they The deposits are located in Alberta and Saskatchewan. Heavy
are relatively thin and the heat losses to the under- and oils having an in situ viscosity less than about 10,000 Mpa.s can
overburden are excessive. To overcome this difficulty, a new be produced by conventional means but only with low
approach has been proposed which utilizes the injection of recoveries. Both bitumens and conventional heavy oils are
saturated hydrocarbon vapours at approximately reservoir produced with higher yields by thermal methods or, in case of
temperature. There are several modifications of the basic shallow tar sand deposits, by surface mining. Thermal methods
principle under development in which the amount of extraneous usually involve the intermittent injection of steam into the
heat injected varies from none to moderate, and the vapour reservoir followed by soaking and production (cyclic steam
chamber is maintained at reservoir temperature to 30C above it. injection), or the continuous injection of steam (steam drive,
This flexibility allows the approach to be followed for the SAGD). Many of the heavy oil reservoirs are relatively thin
recovery of both conventional heavy oils and for bitumen. and the resulting heat losses into the over- and underburden
make steam-based recovery processes uneconomic.
The method, termed "Vapex", utilizes two horizontal wells and
is closely related to the SAGD process but with the steam Recent advances in horizontal drilling techniques make the
chamber replaced by a chamber containing hydrocarbon vapour drilling cost of horizontal wells per unit of total length
near its dew point. comparable to that of vertical wells and horizontal wells are
being drilled extensively in different parts of the world.
The quality of bitumen or heavy oil can be improved Recovery from reservoirs that used to be unattractive with
substantially by in situ deasphalting with propane vapour. The vertical wells has become quite practical with long horizontal
removal of asphaltenes reduces the viscosity of the produced oil wells.
and creates an oil that is more easily refined.
The use of horizontal wells for the non-thermal recovery of
In this paper we present new experimental results obtained in a heavy oils, but at low recoveries, has been economically
scaled physical laboratory model and discuss the effect of factors successful, particularly in Saskatchewan. Horizontal wells have
such as injected Gto ratio, different saturation pressures and also made the SAGD process practical and economically
vapour chamber temperatures on the rate of oil production, the attractive.
extent of in situ deasphalting and the quality of the recovered oil.
The major obstacle to the development of heavy oil and bitumen

409
2 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS

reservoirs is that, because they contain mostly high boiling EXPERIMENTAL WORK
components with about 15 to 20 % by weight of asphaltenes, they Dry Vapex: Apparatus development
have a lower intrinsic value to refiners than conventional crude
oils. The asphaltenes are extremely difficult to convert and they The laboratory research utilized a scaled, two-dimensional
contain nickel, iron and vanadium which poison cracking physical model. The model was confined in a pressure vessel.
catalysts. The viscous crude is also difficult to handle and A diagram of the apparatus is given in Fig.l. The scaled model
requires dilution with light components, or conversion, for cell was made of reinforced phenolic resin sheets and was
transport in pipelines. packed with 1 mm glass beads. It represents a vertical cross-
section through a reservoir with horizontal wells drilled at the
The work described in this paper is directed towards the top and at the bottom of the pay zone. The wells were staggered
development of a low heat requirement technique that can to achieve a better sweep in the initial gas displacement phase of
recover heavy oils and bitumen effectively and that can also the experiment and along the oil-propane interface in the
result in a partial upgrading of heavy crudes in situ so as to drainage phase of the experiment. The cell was equipped with
produce anew, more valuable deasphalted oil that is lighter and 62 thermocouples to obtain temperature readings during the
of a better quality. The undesirable asphaltenes can be left spread of the saturated vapour chamber. The cell temperatures
behind deposited within the reservoir matrix and the recovery were obtained by a Taurus data acquisition system, displayed on
method is environmentally friendly, i.e. it eliminates the need a monitor at predetermined time intervals and recorded on a file.
for large expenditures incurred in clean-up operations that can The thermocouples were calibrated against an external standard
add substantially to the project cost. so that the temperature readings were accurate to within one
degree Celsius. The propane injection vessel was weighed
The method, termed 'Vapex " utilizes pairs of horizontal wells continuously by a load cell and the weight was displayed and
and is closely related to the Steam-Assisted Gravity Drainage recorded.
process[I,2,31. However, in Vapex the steam chamber is replaced
with a chamber containing a hydrocarbon vapour near its dew To maintain the propane pressure at a set value for the duration
point. of the experiment, the propane cylinder was immersed in a
constant temperature water bath; the produced propane was
Although other solvents or their combinations are possible, collected above water in inverted cylinders as free gas and
experiments have shown that the simplest candidate for this solution gas, respectively. The rate of propane injection was not
vapour extraction process is propane. Canada currently controlled directly but was set indirectly by the rate of
produces about 50 million barrels per year of almost pure withdrawal of oil together with its dissolved propane. In tum,
propane (technically Liquefied Petroleum Gas or LPG). Eighty this was determined by the rate of oil drainage from the cell.
percent of the LPG is produced from Canadian natural gas, the Produced oil was collected and its viscosity measured after the
rest is a by-product of crude oil refining. The output of propane removal of propane. Experimental parameters and properties are
is expected to grow by about 6 million barrels by 1995 as U.S. summarized in Table 1.
demand for natural gas increases.
Table 1. Data on the cell, pressure vessel and oil
The problem of extraction, and sometimes in situ deasphalting,
of oil with propane was approached from several directions. Cell material Phenolic resin
Scaled model laboratory experiments were performed in which Cell inside dimensions(LxHxD) 69.8x21.7x3.5 cm
propane alone was injected into the cell (Dry Vapex) both under Number of I-type thermocouples 62
steady state conditions, and under a pressure cycling regime. In Packing 1 mm glass beads
another set of experiments, propane was injected with steam Packing density 1.41 kg/litre
(Wet Vapex) to study the coexistence of a large, low temperature Cell permeability 81030 ILrrt-
propane chamber with a hot steam chamber limited to the Porosity 0.391
proximity of the injector/producer. In separate mixing tests Pressure vessel: max.operating pressure 8.3 Mpa
propane was allowed to contact different amounts of oil and the typical operating pressure 0.9 - 1 MPa
resulting deasphalting was related to the propane-to-oil (P/O) Oil type North Tangleflags (Lloydminster)
ratio in the apparatus. The samples of deasphalted oil (DAO) Oil viscosity 10,000 mPa.s at 200C
were used for analyses of trace metals (V, Ni, Fe) to assess the Oil density 0.979 g/ml at 15.56C
suitability of the oil for rerming. API gravity 13.0
Asphaltene content 15.6 wt. % (cold pentane)
Initial cell temperature 20C

410
 I.l. MOKRYS and R.M. BUTLER 3

RESULTS OF DRY VAPEX RUNS: Run 3.3 (Fig. 5) represents an attractive trade-off between insitu
(a) Steady-State Conditions upgrading and the rate of production for a steady state operation.
The viscosity of produced oil reached 1000 mPa.s at 25%
A total of five experiments was performed at room temperature production and kept decreasing until the experiment was stopped
and constant cell pressures, covering a range of 708-984 kPa, ie. at 50% production. The drainage rate was essentially the same
from below the propane dew point to pressures above it (Fig.2). as for run 3.2. The opened cell showed a large consolidated
The experiments were designed to show the effect of propane area of beads where the asphaltenes from the oil were deposited.
equilibrium pressure on the oil drainage rate, percent recovery, The produced area of the cell is about 75 % of the total area
asphaltene deposition in the cell and the produced oil viscosity. while the amount of oil produced and weighed is only 50 %.
The difference is due to asphaltenes that remained in the cell.
For all experiments the cell was initially at 20C; the propane
dew point at this temperature is 750 kPa (average Pbar =87.9 (b) Pressure cycling
kPa). During the course of each experiment, as the propane
dissolves in the freshly exposed surface of the draining oil at the It was thought that operation with the cycling of pressure might
propane-oil interface, the heat of condensation of propane creates promote both mixing and the formation of asphaltene-free flow
a region of raised temperature (a heat front) that spreads through paths for the diluted bitumen. To put this idea to a test
the cell, away from the plane of injector/producer, as the oil experiments were carried out in which the propane pressure was
drains and the chamber grows. Although this increase in maintained for the first 20 minutes at some predetermined value,
temperature is typically only 4-6C, it has a local effect on the then allowed to fall, by shutting off the propane supply, to a
propane-oil equilibrium. value 34 kPa lower, and 20 minutes later the pressure was
slowly built up to its initial value. The propane pressure cycle
The results of Fig.2 show that the cumulative oil production is illustrated in Fig.6 and the results of the runs are summarized
after 9 hours of experimental time peaks between 846-915 kPa in Fig. 7. The data show a maximum production at about
under steady-state conditions. In the undersaturated region of 846 kPa propane pressure.
the plot the rate and the oil production are lower than the
maximum because of decreased solubility of propane vapour Final stabilized drainage rates in both the laboratory and field
below the dew point. In the region of oversaturation the units for steady state and pressure cycling conditions are
propane vapour condenses into the oil surface in quantities compared in Fig.8. The scaling to field units[4) assumes a 10m
exceeding the limit of about 35 % propane by wt., which causes thick payzone, 457 m long horizontal wells and drainage
asphaltene precipitation. If the propane pressure is raised above occurring from both sides of the well. The plots show that the
915 kPa at steady-state conditions, the asphaUene particle pressure cycling of propane results in an increase of the final
precipitation on the surface of the beads occurs en masse stabilized drainage rate of up to 40 % over the steady state
resulting in decreased permeability of the consolidated drained method. The steady state method, on the other hand, results in
area. The drainage flow of deasphalted oil is then impaired and an overall 10% larger cumulative oil recovery after 9 hours of
the production rate lowered. model time (4.9 years in the field), because the higher initial gas
displacement rate offset the lower final stabilized drainage rate.
The comparison is apparent from runs 3.1-3.3 in Figs. 3, 4 and
5. In run 3.1 the propane vapour was undersaturated, producing A comparison of Fig.7 for cycling conditions with Fig.2 for
40 % of the OOIP in nine model hours, or 5 field years, with a steady state conditions reveals that cumulative oil production
final field stabilized rate of 98 bbl/d (10m payzone, 1500ft H- depends less on the propane vapour saturation in the pressure
well, 2 sides). Although the results are scattered, the produced cycling tests, particularly in the early stages of the run. For
oil viscosity was essentially unchanged. The photo shows the example, after 5 hours of model time the cumulative production
opened cell after the run. The clean yellow extracted beads for the steady-state experiments exhibits a sharp peak at about
show no evidence of asphaltene deposition and a sharp oil 881 kPa, whereas the corresponding oil production curve for
drainage interface was visible. pressure cycling is fairly flat and shows little change between
777-984 kPa. Even at 9 hours of model time the pressure
Run 3.2 (Fig. 4) was carried out with propane vapour saturated cycling oil production is flatter and peaks at about 64 kPa lower
at the original reservoir temperature. This resulted in 115 bblld pressure (812 kPa) than the steady-state curve (881 kPa).
stabilized production rate and a steady decrease in viscosity from
the original 10,000 mPa.s to about 4,000 mPa.s. The The practical implication for the production from an oil well
photograph of the opened cell shows bands of asphaltenes which using this extraction method is that it would appear that pressure
were deposited as the interface receded during drainage. cycling could be advantageous after the initial displacement
phase. More experimentation is required to confirm this.

411
4 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS

Steam-propane injection (Wet Vapex) A comparison of the steam (3.15) and steam/propane (3.16) runs
is given in Table 2.
Simultaneous steam-propane injection into the scaled model is a
variation on the Steam-Assisted Gravity Drainage process with Table 2 Summary of steam and steam/propane runs
propane forming a low temperature oil production zone that
spreads laterally away from the steam zone at the Steam Steam/propane
injector/producer. This approach results in potential savings on
the cost of injecting steam and is suitable for reservoirs where 1 mm glass beads 7.099kg 7.099kg
heat losses are high. The steam is used to create a limited hot Packing permeability 830 darcy 830 darcy
region (ie. a steam chamber) in the vicinity of the Tangleflags heavy oil 1.792kg 1. 837kg
injector/producer in which the propane is stripped from the Oil viscosity at 200C 10,OOOm Pa.s 10,OOOm Pa.s
draining oil and recycled into the laterally spreading, cooler Steam pressure 1088 kPa 1108 kPa
propane chamber. As a result of this internal recycling Steam temperature 181C 185C
mechanism the G/O ratios are low and the (latent) heat of Propane pressure 148 psig
propane vapour is transported to the oil-propane interface that is Propane saturation temp. 31C
receding deeper into the cold reservoir. Steam consumptionlhr 1.10 kg 0.76 kg
Heat inputlhr 3074 kJ 2118kJ
The experimental conditions for the recovery of a Lloydminster Oil recovery after 6 hrs 30.9% 27.0%
type oil from the scaled model were chosen such that both Total propane consumption - 0.102 kg
chambers were maintained at 1108 kPa, making the saturation Propane recovery 99%
temperature for propane 31C and for steam 185C,
respectively. The scaled physical model contained 1 mm glass The heat distribution in the cell near the end of the run (at about
beads as a porous medium and was filled with Tangleflags heavy 7 hrs.) is illustrated in Fig. 12. The figure contains
oil. thermocouple temperature readings and lines of constant
saturation for propane. The positions of these lines were
The modified Vapex apparatus is illustrated in Fig.9. The basic estimated from the vapour pressure of propane-oil mixtures
apparatus was expanded to include a steam generator/pressure assuming ideal behaviour (Fig.13) and the calculation for the
regulator, propane pump/pulse dampener and a steam condensate construction of the graph was discussed elsewherel51 The
knock-out pot. An in-line regulator was used to supply constant temperature readings indicate that the hot steam zone is limited
pressure steam to the scaled physical model. The steam line to a narrow region around the injector/producer while the cooler
passed, via a specially insulated connection, through the thick propane chamber spread laterally to the far end of the cell,
flange of the pressure vessel which, even with the insulation, replacing the drained diluted oil.
acted as a heat sink and caused some condensation of steam.
This was overcome by adding a condensate knock-out pot inside At low propane saturation the lines follow the shape of the rising
the pressure vessel. Most of the steam from the knock-out pot steam chamber. At 50% w/w propane in the propane-oil
was returned directly back through the flange and only a mixture, the propane saturation lines have assumed the shape of
sidestream was used for the process itself. This sidestream was the observed drainage interface as visible on the photograph of
superheated slightly before introduction to the cell. The distance the open cell at the end of the experiment (Fig. 12). Since, as a
between the center of the injection and production port in the result of the release of latent heat of solution of propane, the oil-
scaled model was 7/8 inch (2.22 cm). propane interface has warmed up by about 5C, the observed
drainage interface is located above the one calculated (saturated
Prior to the start of the experiment steam was passed through the C3 line) and roughly coincides with the 50%~(wt) saturation
warm-up section of the injector/producer (Fig. 10) for about 15 line. The 100% saturation line is only hypothetical since it falls
minutes to create a communication path for the initial drainage. beyond the gas/liquid interface.
Dry steam was then injected into the cell for the first 75 minutes
of the experiment. The rate of steam injection averaged about Fig.ll shows the cumulative oil production, water-to-oil and
0.77 kg/hr and of propane 12 g/hr throughout the run. The gas-to-oil ratios for the steam-propane experiment. For
solution propane gas was separated, measured and the collected comparison, a recovery curve is shown from a parallel
liquid samples (which included water-oil emulsions) were experiment in which steam alone was injected into the cell (with
analyzed for water by the Karl-Fischer method. The amount of the same oil, beads and injector/producer configuration) under
recovered oil was then back-calculated. almost identical conditions (1088 kPa at 181C). Also in Fig.Il
is a photograph of the opened cell at the end of the run. Pieces

412
 I.J. MOKRYS and R.M. BUTLER 5

of white string were added to define the position of the drainage at 200C and with a propane-to-oil ratio of 3 (w/w). The
interface. viscosity curves exhibit a threshold value which the propane
concentration must reach before the onset of deasphalting.
The scaled model results show that after 3 years of field time the
steam-propane production of oil is comparable to that obtained For a comparison, Figs. 14 and 15 also contain the viscosities of
by the SAGO method. The steam consumption for the propane- pentane deasphalted oil (80: I excess at 20C); the viscosity of
steam process was 69 % of the steam-only process in the two- the pentane deasphalted oil is ten times higher than that of
dimensional laboratory model; this decrease could become more propane deasphalted oil.
pronounced in the field.
The chemical natures of these precipitated asphaltenes are
Energy Considerations different. Whereas excess propane precipitates both high and
medium molecular mass asphaltenes that usually form a very
The dry Vapex process is inherently more energy efficient than viscous sticky material, pentane separates only a high molecular
SAGO since typically only 0.2-0.5 kg of propane is required for mass powdery substance.
a produr.tion of I kg of oil as opposed to 3 kg of steam/kg of
oil. In addition, the latent heat of propane is only about 1/6 that In our earlier work with the packed scaled model it was
of water; the initial reservoir temperature is raised by no more sometimes observed that the viscosity of the DAO increased
than 5-10oC, as opposed to about 200C for the steam process. above its initial value towards the end of the run. Although
The dry propane process does not subject the reservoir to high these occurrences were rare, a question arose as to whether
temperature steam and there is no danger of reservoir damage under some conditions asphaltenes could re-dissolve in the oil.
from the action of steam condensate on sensitive clays. The use A test was therefore performed in which pentane precipitated
of propane eliminates the need for a fresh water supply and for asphaltenes were added in small amounts to a sample of the
a costly environmental clean-up or water treatment and recycling mother DAO, mixed at 20C for 2-4 hours and the viscosity was
facility. determined after each stepwise addition. The results are plotted
in Fig. 16. They show that the process is reversible and that by
Bench-top deasphalting of bitumen and heavy oil with propane adding asphaltenes to DAO it is not only possible to reconstitute
and pentane the original bitumen but, by adding more than the precipitated
amount of asphaltenes, to create a new crude of higher viscosity
. Cold Lake bitumen and Lloydminster heavy oil were deasphalted than the original bitumen .
with propane as a precipitant. The deasphalting was carried out
in a pressure cylinder packed with a lmm glass beads; these The metal removal resulting from deasphalting is illustrated in
served as a substrate for the deposition of the precipitate. The Fig. 17 . The results show that the total content of V, Ni and Fe
contents were mixed by tumbling the cylinder end over end at (total metal content) decreases from 270 ppm to 10 ppm for
regular times for a period of four hours. Cold Lake bitumen. A similar set of results for the
Lloydminster oil shows V, Ni and Fe diminish from 185 ppm to
The viscosity of the dead deasphalted oil with increasing 13 ppm. Since these metals are active catalyst poisons for
amounts of propane in the oil-propane mixture is plotted in processes such as catalytic cracking and hydrocracking, their
Fig. 14 for Cold Lake bitumen and Fig. 15 for the Lloydminster removal through deasphalting is an important aspect of the
oil. Cold Lake bitumen properties are listed in Table 3 and Vapex in situ upgrading. It should be much cheaper to upgrade
those for Lloydminster oil are in Table 1. these deasphalted oils than the originals.

Table 3. Cold Lake bitumen CONCLUSIONS

Viscosity 100,000 mPa.s at 20C 1. The rate of Vapex extraction of heavy oils with saturated
Density 0.9955 g/cm3 at 15.56C propane under steady state conditions is largest at 846-915 kPa
API gravity 10.6 for an initial reservoir temperature of 200C. At higher propane
Asphaltene content 17.1 wt% (cold pentane) pressures asphaltenes are precipitated en masse causing
impairment of drainage flow. Stabilized rates of the order of
During deasphalting the pressure was above the vapour pressure 206 bbl/d (33 m3 /d) are possible with Lloydminster type oils
of propane at 200C and there was no vapour phase. from a 10m pay zone and a 1000m long horizontal well.

The results in Figs. 14 and 15 indicate a viscosity decrease by a 2. With pressure cycling, the saturated propane vapour injection
factor of 300 for Cold Lake bitumen and 50 for Lloydminster oil into a 10m thick heavy oil reservoir results in final, stabilized

413
6 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS

drainage rates of about 348 bblld (55 m3 /d) using a l000m 3. Butler, R.M. "Thermal Recovery of Oil and Bitumen",
horizontal producer. Englewood Cliffs: N.J., Prentice Hall (1991).

3. Pressure cycling could be advantageous after the initial 4. Butler, R.M. and Mokrys, LJ. "A New Process (Vapex) for
displacement phase of the process. Recovering Heavy Oils Using Hot Water and Hydrocarbon
Vapour", J. Can. Pet. Tech., Vo1.30, No.1, pp. 97-106, Jan.-
4. The propane recovered oil can be upgraded in situ and thus Feb. (1991).
be of higher quality than the original oil. The viscosity of
Lloydminster type heavy oil was lowered by a factor as high as 5. Butler, R.M. and Mokrys, LJ. "Recovery of Heavy Oils
50 in scaled model recovery experiments. Using Vapourized Hydrocarbon Solvents: Further Development
of the Vapex Process". Presented at the 4th Petroleum
5. Simultaneous steam-propane injection results in a heavy oil Conference of the S. Saskatchewan Section, Petroleum Society
production that is comparable to that of steam injection. The of CIM, Regina, Oct.7-9, 1991. Accepted for publication in the
steam-propane process is inherently more energy efficient than J. Can. Pet. Tech. (In press).
SAGD. The steam is used to create a limited hot region in
which the propane is stripped from the draining oil and recycled
internally. The cooler propane spreads laterally into the
reservoir where it dilutes, upgrades and recovers oil.

6. Benchtop mixing experiments have shown that the viscosity

of propane upgraded crude can be lowered by a factor of 50 and
300 for the Lloydminster heavy oil and Cold Lake bitumen,
respectively. The oil can be reconstituted from asphaltenes and
maltenes and the most effective temperature for deasphalting is
in the range of 10-300C.

7. The total content of V,Ni and Fe in the deasphalted Cold

Lake bitumen decreased from 270 ppm to 10 ppm in bench
deasphalting experiments. These changes would make the DAO
more amenable to direct refining.

ACKNOWLEDGEMENTS

The authors wish to acknowledge the contributions of Ronald

Turner and Clarence Urness who carried out the experimental
work and modified the apparatus. This work was carried out
under a joint research contract with the Canadian Centre for
Mineral & Energy Technology (CANMEn, Energy Mines and
Resources, Canada and a consortium of six oil companies -
Imperial Oil Resources, Shell Canada, Amoco, Saskoil, Sceptre
Resources and PanCanadian Petroleum. The authors are grateful
for this financial support.

REFERENCES

1. Butler, R.M., McNab, G.S. and Lo, H.Y. "Theoretical

Studies on the Gravity Drainage of Heavy Oil During In Situ
Steam Heating", Can. J. Chern. Eng., 59, pp. 455-460, (1981).

2. Butler, R.M. and Stephens, D.J. "The Gravity Drainage of

Steam Heated Heavy Oil to Parallel Horizontal Wells", J. Can.
Pet. Tech., pp. 90-96, April-June (1981).

414
 1.]. MOKRYS and R.M. BUTLER 7

COMPUTER
(Temp.welght-nne)

FREE GAS SOLUTION GAS

TAURUS

PRESSURE VESSEL

FIG.1
VAPEX: SCHEMATIC OF PROPANE INJECTION APPARATUS

_1,~P-S~---Y-S-T~A~TE~C~O-N-D-IT-IO-~~----------------~
.9 PARAMETER IS EXPERIMENTAL
~ 900 TIME IN (hrs)
i= 20C. DEW PT. =750 kPa(ga
!:<
-J 800 P ,,; 88 kPa
ba
o
~ 700
0..
....J
o 600
~ 8
~500 7
6
::> 400 5
:E
::> 4
()3QO
3
200 2
100 1
UNDERSATURATION OVERSATURATION

o~--------------------------------~
483 552 621 689 758 827 896 965
CELL EQUIUBRIUM PRESSURE (kPa)
FIG.2 STEADY-STATE OIL PRODUCTION
415
8 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS
FIELD TIME (yrs)
o 1 2 3 4 5
1000~------~------~------~------~------~-' z
- . Tangleflags heavy oil
50
.9
Z
o
900 1 mm glass beads
Undersaturated propane only
P=100 ~ig
o
:::>
I- 800 . T=22-2S'C Drainage rate:
o
U 56g/hr::OObbl/d
(10m payzone,457m H-well)
40~
5700 a.
o Displacement rate: I-
z
8: 600 426g/hr-74Obblfd
(10m payzone,457m H-well) 30~
..J ~~ 12000 a:
0500 8:a lii-10000
w
w a.
>400
II)

5 % 8000 .
I-
(Dead oil) 20
G/O(lnjctd)
:5 300
:::> ~2
~ ~
~1
u:::> 200 :E
G/O(recvd) - 10
::::l
100 1 Model hour =0.54 Field years 0 00 10 20 30 40
PERCENT OIL PRODUCTION

o0 o
1 2 3 4 5 6 7 8 9 10
MODEL TIME (hrs)

FIG.3 RESULTS OF VAPEX RUN 3.1

416
 U. MOKRYS and R.M. BUTLER 9

FIELD TIME (hrs)

0 1 2 3 4 5
1,000 HEAVY OIL
...-... TANGLEFLAGS (LLOYDMINSTE
-9 900 1 mm glass beads - 50>-
Propane only c:
Z Cell P=110 pslg W
0 Chamber temp. =23-28 deg.C >
r- 800 r- ....
0
- 400
0 W
Initial rate:
::::> 700 , c:
0 473g!hr-821 bbl/d
0 (10m payzone,457m H-well)
~" r-
ex: Viscosity of native 011 Z
a.. 600
E,
eg- 30~
.....J
0 500 Final rate:
~ ... C\I
'lii
c:
W
~7 a..
W 66g!hr :::::115bbl/d
80 ~
>400 i0 20
~
~300
::::>
, (j')
:>
....J
0
la
w
0
~
~200 ~
0
G/O recvd)
:5
~ 10
0 o~
~

0
100 '0 20 30 40 50 110

1 Model hour = 0.54 Field years PERCENT OIL PRODUCTION

0 0
0 1 2 3 4 5 6 7 8 9 10
MODEL TIME (hrs)

FIG.4 RESULTS OF VAPEX RUN 3.2

417
10 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING:
THE VAPEX PROCESS
FIELD TIME (yrs)
1 2 3 4 5
1,000 0

-
,9900
z
TANGLEFLAGS (LLOYDMINSTER) HEAVY OIL
1 mm glass beads
Propane only
Cell P=120 psig
50
>-
a:
W
0 800
Chamber temp. =23-30 deg.C
>
r- Drainage rate: ~ 40 0
0 53g/hr :::94bbl/d 11000
0
::> 700 Native oil viscosity W
0
(10m payzone,457m H-welQ ~10000 a:
0 E g(l()() r-
a: 600 <5 Z
a. Displacement rate: 0 8000
N
30W
...J 387g/hr ::::684bbl/d 1ii 7000 0
0 500 (10m payzone, ~eooo
a:
W
W 457m H-well)
~sooo a.
> 400 ~ ()
4000

,
C/)

5 :> fii' 20

::> 300
...J 3000

c2000
i
0
~
::>
Lfi 1000
c 0 a-
200 1~
0 10
~
~
0.5:5
GIO(racvd) ::::>
100 - ~
1 Model hour = .54 Field years 0::::>
()

0 0
0 1 2 3 4 5 6 7 8 9 10
MODEL TIME (hrs)

FIG.5 RESULTS OF VAPEX RUN 3.3

418
 I.J. MOKRYS and R.M. BUTLER 11

~ Dew point pressure

Max.
press.
;~re~surepres,ure . .
depletion i build-U

o 20
-.----------- .60 t (mins) ONECYC~
40

FIG.6 PROPANE PRESSURE CYCLING

1.000m~nlnmm~mr.mmm!l"Re~
PARAMETER IS EXPERIMENTAL TIME (hre)
S GOO DEW PT.- 748.8 kPa uge)
~ t,.- 87.81 kPa

~:
oJ eoo
o
W 500

~:::>
!~r
400

a
~
300 3~ j_
0 :

200 2 : . .
RUN3.13
""--RUN3.10 FiJN3.7
100 1 -'RJN3. 12 . . : RliN3.11
UNDERSATURATlON: OVERSATURATlON

o~--------~----------------~
552 821 889 758 827 865
CaL MAXIMUM PRESSURE (!cPa)

FIG.7 PRODUCTION AS A FUNCTION OF PROPANE SATURATION

~r-------------------------~
TANGLEGL.AGS HEAVY OIL
SINcIy et.m. conditions

o Pressure cycling
Aaphaitlne drainage impairment

;g 1SO

-
:8
w Vapour chamber}
21/2

iw 100
temperature
atart/end

~
Z
~
o 3.8@
221215
so Undereaturated Oversaturated
propane t-\ propane
3.15 vapour ~ vapour
3.14
2G'24
C.llinitially at 2O-2f C. dew point 758 kPa

o~--------~----------------
552 621 689 758 827 896 ..965
CELL EQUIUBRIUM (OR MAX.) PRESSURE (kPa)
FIG.8 VAPEX: OIL PRODUCTION AT DIFFERENT SATURATIONS
419
12 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS

--
NOT TO SCALE

COMPUTER
(T.weig ht,rate) SOLUTlON Gf.8

- F1g.10 .---TA-U......R-U-S---,

SAFETY
BABOON

GENERATOR

VAPEX: SCHEMATIC OF PROPANE-STEAM INJECTION APPARATUS

FIG.9

. . . . . . . . . .
...........
.................. ..
. . . ...
. .
....... ... .
. ...
.... ............ '
.
. . ......
. . . . ................ ...
. . .... .. . .... .
. . . . . . . . . . . . . . ........ ...........
.....
. . . ....
....
...
.
... ...... .......
......... ...... ..... ........ . . ........ .
......... '
.... ... .......
'
, .
. .........
... ............ . ..
...... ..... ...........
. ........ . ........... . '.
. . . . . . . . .. . ...................
"',
...... -.......... ','.
. ..................... .
. ............... .......
,', ........... .
"

......... ............. . . ... . .........

. . . . . . . . .. .
.. . . . ..."
. . . ..
.. . .. . .. . . " "
"

..... '::::.:::::::: .:- -:.:-:-: :-:-:-: :::: ::>:-'." . GLASS

..
...... ........ . ...... ............ . :< :.;...---
......... .....
BEADS
SATURATED
WITH
BITUMEN BITUMEN
FILL-UP
PORT

INJECTION PORT
FOR STEAM AND
\':\.)/%$1-- O-RING

~ iiiliilli:= (START
PROPANE
(OR WARM-UP OUTLET UP INLET
WARM-UP WITH STEAM)
PORT
PORT) PRODUCTION PORT
(BITUMEN/SOLN. GAS)

VAPEX: HEATED INJECTOR/PRODUCER

FIG.10
420
 1.1. MOKRYS and R.M. BUTLER 13

FIELD TIME (yrs)

700pO______~1------~2-------3~----~4----~
-
S
z600
TANGLEFLAGS (LLOYDMINSTER) OIL
STEAM-PROPANE INJECTION
STEAM PRESSURE = 148 PSIG
Steam only
(920413)
87bbl/d
o TEMPERATURE = 185 DEG.C
30>-
t= 1 mm GLASS BEADS a:
gsoo INJ-PROD 7/8 OF AN INCH APART. W
c 33.7g/hr= 6Obbl/d
~
~400
(10m payzone, 1 Model hour = 0.54 Field years ()
457m H-well)
a..
~ 15 20~
--' 10 r-
5300
5 ICUM.PROD.w/O (MASS) I zW
W ()
> Or-------------~

-200
~
0.9
0.6 OWa:a.
0.3
~100
:::> 10 20 30
() PERCENT OIL PRODUCTION

O~----------------------~-----------------O
o 1 2 3 4 567 8 9
MODEL TIME (hrs)

FIG.11 RESULTS OF STEAM-PROPANE RUN 3.16

421
14 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS
DRAWN TO SCALE
STEAM ZONE UNDERSATURATED PROPANE
2O%C3 (wt)

SATURATED C
s
32

INJECTOR
& OBSERVED DRAINAGE INTERFACE
PRODUCER
2D VAPEX CELL DIMENSIONS: RUN 3.16 CONDITIONS:
INSIDE HEIGHT = 21.70 em
INSIDE WIDTH = 69.80 em F!..m= 1020 kPa(gauge) =
T.INm 185C
CELL VOLUME!::: 5.150 J Pprop = 1020 kPa(gauge) Tprop = 31C
OIL VOLUME!::: 1.81 J

FIG.12 RUN 3.16: TEMPERATURE DISTRIBUTION AT 6:54:04 HRS

90 80 70 80 50 40 30 20 TOe

PARAMETER IS WEIGHT FRACTION

500 OF OIL IN THE OIL-PROPANE 3,447
IDEAL SOLUTION
300

.m 200
IU
Cl.
!. 1.379
~

ui 150 ui
a: a:
::> 0 ::>
en 100 .5 690 en
en
UJ .8 en
UJ
a: a:
a.. .9 Cl.
Z Z
0 50 345 0
F= .95 F=
::> ::>
...J 30 ...J
0 0
en en
20 138
15

.99
10 69

2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5

~ *1000, K-1
FIG.13 PROPANE-OIL COMPOSITION-P-T DIAGRAM
422
 I.I. MOKRYS and R.M. BUTLER 15

0102030 40 50 60 65 70 wt.% C3

COLD LAKE BITUMEN

100,000 Native bitumen 17.1
en
rP
o 15~cP
~ 30,000 ~o
III Ci!~
aI FO.ntane 10 ifen ~
~ 10,000
deuphalted
<
!II
J
(5
(e:Jr--2Q
bItUmen 0
C) 5 '# ~
"E

~ 3,000 :.:.:.:.:.:.:.:.:.:.:.:.:.:-:.:.:.:.::. ........................ :.:.:-:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:-:.:-:.:.:.:.:.:.:.:.:.:.:.:: 0 ~

o
~
~ 1,000 .

8
CIJ
:;
20'b
300

1ooW.------------------~~~
o 0.5 1 1.5 2 2.5 3
PROPANE/OIL RATIO (Mass)
FIG.14 PROPANE DEASPHALTING:
EFFECT OF PROPANE/OIL RATIO

20,000 0 2030 40 50 60 65 70 wt.% C3

() LLOYDMINSTER HEAVY OIL ru
z
c:oN 10, Native oil
w_
5~ ()
<
i 5,000 Pentane
OI
<0
3; or
N

a..
-o
E
. .J
;;
deasphaJted oil
(80:1 wtexoeaa,2OC)
5
*'< If.~
::.':":'::'::'::::.:-:-:.::.:.:.:.:.:.:.:.:.:.:.:..:.:.:.: .:.:.:.:.: . . .:.:.:.:.:.::::.:.:.:.::.:.::.:.:::.::.:' 0 :a:;:
2,000 :>-

o~ 1,000
LL
o 500
~
{jj
o()
en 200
>
100 0 0.5 1 1.5 2 2.5 3
PROPANE/OIL RATIO (Mass)
FIG.15 PROPANE DEASPHALTING:
EFFECT OF PROPANE/OIL RATIO
423
16 IN SITU UPGRADING OF HEAVY OILS AND BITUMEN BY PROPANE DEASPHALTING: SPE 25452
THE VAPEX PROCESS
1 , 0 0 0 , 0 0 0 , . - - - - - - - - - - - - - - - - - -...
PEACE RIVER BITUMEN
~ 500,000 18.8-19.5% ASPHALTENES (70 to 80:1 Pentane at 20 deg.C)
(l. MIXING nME 24 HAS AT 20 DEG.C

E~ 200,000
(,) Native viscosity = 123,500 mPas at20 deg.C
0 100,000
C\I
~ 50,000

~
(J)
20,000
o(,) 10,000
(J)
:> 5,000
zw
~ 2,000 '-- PENTANE DEASPHALTED BITUMEN
::>
~ 1,00
co 0 2 4 6 8 1 12 14 16 18 20
ASPHALTENE CONTENT (wt.%)
FIG.16 VISCOSITY OF RE-CONSTITUTED BITUMEN SAMPLES

COLD LAKE BITUMEN

Propane Deasphalted Samples

DEASPHALTED
BITUMEN:
Total
V,Ni,Fe
Total

,
V,Ni,Fe
270.4 ppm
10.2 ppm

424