Note: Descriptions are shown in the official language in which they were submitted.
CA 02374115 2002-03-01
1
Title: Energy Efficient Method and Apparatus for Stimulating
Heavy Oil Production
FIELD OF THE INVENTION
This invention relates generally to the field of enhanced
resource recovery, and more particularly to enhanced recovery
of heavy oils and bitumen. Most particularly this invention
relates to energy utilization in such resource recovery.
BACKGROUND OF THE INVENTION
Currently steam is the dominant thermal fluid used for
insitu recovery of bitumen and heavy oil. Steam raises the
temperature of the bitumen thereby reducing its viscosity and
allowing it to flow more easily. Steam extraction is subject to a
number of problems including high heat losses, clay swelling
problems, thief zones, water-oil emulsions, capillary surface
tension effects, lack of confinement for shallower zones and
disposal of large quantities of environmentally damaging salt and
organic acids as a consequence of boiler feed water purity
requirements. With the best available technologies, only 10% of
the original bitumen resource is economic to extract.
Thermal recovery processes, using steam, require large
amounts of fuel to be burned to produce the steam and can emit
enormous amounts of greenhouse gases such as carbon dioxide.
Estimates published by Natural Resources Canada' show C02
emissions of about 70kg/bbl for bitumen production and a total of
about 120 kg/bbl for synthetic crude (i.e. upgraded bitumen
usually derived from surface mined bitumen).
Alternative sources of steam such as co-generation may
offer some synergies with lower "unit" emissions (per kWH or per
kg steam). However, co-generation also produces higher gross
Canada's Emissions Outlook: an Update, December 1999,
Annex B, pg B-6, Available at www.rirCai'.gc.ca?esiC:eo/update.htt3't
CA 02374115 2002-03-01
2
emissions than direct steam generation, due to the fact that
additional fuel is burnt to generate electricity. From an
environmental perspective, higher total emissions means more
environmental impact (even if the unit emissions are reduced).
Recent estimates released by the Alberta Energy Utilities
Board2 and the Canadian Association of Petroleum Producers3,
predict that bitumen (and synthetic crude) production rates will be
2 to 2.6 million bbl/day of bitumen by 2010. This level of bitumen
production will produce at least 140 million kilograms
(=70x2million) of C02 emissions per day (i.e. 300,000,000 to
700,000,000lbs C02 per day depending on the proportion of
insitu vs synthetic crude production).
With the recent spike in natural gas prices many oil
companies are looking for altemative fuels which are less costly.
A recent patent application 2,332,685 proposes to deasphalt the
SAGD bitumen in surface facilities and then bum the asphaltene
residue to generate steam. This patent application teaches
substitution of a clean buming fuel (natural gas) for a less
expensive asphaltene residue which is high in carbon, sulphur
(>8%) and toxic metals. 2,332,685 would double the C02
emissions for SAGD bitumen production as compared to a
process burning natural gas. While this process may have
economic benefits, the benefits come at a cost of substantially
higher emissions into the environment.
What is needed therefore a way to reduce the energy
requirement for bitumen extraction.
BRIEF SUMMARY OF THE INVENTION
The present invention is directed to a process and
apparatus to reduce energy requirements for the production of
2 Alberta's Reserves 2000 and Supply/Demand outlook 2001-
2010, Alberta Energy Utilities Board
3 Canada's Oil Sands Development delivered by Eric
Newell, Chairman & CEO, Syncrude Canada. Available at
http://www.capp.ca/
CA 02374115 2002-03-01
3
heavy oil and bitumen by solvent extraction. This process also
simultaneously upgrades the heavy oil and bitumen insitu. In one
aspect of the invention a heat pump is used to recover heat from
a latent or free heat source to reduce the energy load required to
prepare the solvent for injection and extraction. In another aspect
heat pumps are used to recover the solvent from the produced
fluids to again reduce direct energy consumption and to reduce
emissions. In another aspect of the invention a production
apparatus is provided which is capable of utilizing the energy
efficient and pollution reducing processes of the present
invention.
As outlined in more detail below, the present invention is
predicted to reduce the amount of energy required for extraction
by 90-95% percent relative to the best prior art insitu technology,
i.e., steam assisted gravity drainage (SAGD). By reducing energy
costs, a primary expense, the present invention also increases in
the value of the production, improving both the operating and
profit margins. Most importantly, the process and apparatus of
the present invention is predicted to reduce the amount of
emissions by 90-95% relative to the prior art SAGD.
The overall energy required by the extraction process is
determined by the need to heat the oil/bitumen reservoir to the
extraction temperature. The present invention is most preferably
used in conjunction with an extraction process in which the
bitumen extraction temperatures are provided by using a
condensing solvent such as propane. By utilizing such a
condensing solvent and the latent heat of condensation, the
energy requirement for extraction can be reduced thereby
reducing consequent emissions over SAGD methods.
For example if the original reservoir temperature is 8C
and the solvent extraction process operates at a temperature of
50C (vs 230C for a steam extraction process) then the expected
energy savings of the solvent extraction process would be 81%
CA 02374115 2002-03-01
4
(=(230-50)/(230-8)) relative to current steam extraction
technology (SAGD). The present invention delivers additional
energy and hence emissions savings over and above this level
as explained in more detail below.
The heat for bitumen extraction is supplied by the latent
heat of condensation of the solvent. The present invention is
comprehends choosing the solvent such as propane, which has
an appropriate boiling point of -42C. In this way much of the
required latent heat of vaporization can be supplied via heat
exchange with "free" or latent sources of heat such as ambient air
or water, or even waste heat from flue gas or the like. The
vaporized solvent is subsequently compressed, thereby raising its
temperature and pressure to the desired extraction conditions.
Thus, the present invention comprehends a hydrocarbon (i.e.
heavy oil) extraction process in which the solvent absorbs most of
its heat energy at low temperature and then a small amount of
compression energy is used to raise the solvent condensation
temperature to the preferred extraction temperature analogous to
a heat pump or a Carnot refrigerator. Thus, for a solvent
extraction process operating at 50C, with absorption of heat from
a latent heat source at 5C according to the present invention (i.e.
an arbitrary but "reasonable" heat source temperature), the
thermodynamic coefficient of performance is 6.2 (=278K/(323K-
278K)). This means that according to the methods and
apparatuses of the present invention one joule of work in the
compressor can deliver 6.2 joules of heat at 50C.
Therefore according to one aspect of the present
invention there is provided a method for extracting oil from a
formation, said method comprising:
Selecting a condensing solvent having a high vapour
pressure at low temperatures,
Injecting said solvent as a vapour into said formation,
Condensing said solvent in said formation to deliver a
CA 02374115 2002-03-01
latent heat of condensation to said oil and to mobilize said oil,
Producing said oil and said solvent;
Recovering said solvent from said produced oil,
Heating said solvent means of a heat pump, and
5 Re-injecting said solvent into said formation.
According to another aspect of the present invention there
is provided a method of stripping a solvent from produced oil and
preparing the solvent for re-injection, said method comprises the
steps of:
De-watering said produced oil;
Separating said produced oil into a gas fraction and
a liquid fraction;
Removing non-condensibles from said gas fraction;
Passing said liquids through one or more
evaporation steps to recover additional solvent; and
Vaporizing said recovered solvent.
According to yet another aspect of the present invention
there is provided an oil recovery system comprising:
a source of solvent for injection into a hydrocarbon
bearing formation;
an injection path into an oil bearing formation;
a recovery path out of said oil bearing formation;
and
a solvent recovery and reheating apparatus
including at least one heat pump.
BRIEF DESCRIPTION OF THE DRAWINGS
Reference will now be made, by way of example only, to
preferred embodiments of the invention as illustrated in the
accompanying drawings and in which:
Fig. 1 shows a process flow diagram for one embodiment
of the present invention including solvent recovery, purification
and solvent vaporization.
CA 02374115 2002-03-01
6
Fig. 2 shows a simplified Molliere diagram for propane,
which identifies two different paths to provide hot, high pressure
propane vapour for insitu bitumen extraction, to illustrate the
thermodynamic energy requirements for the present invention as
compared to a direct fired vaporization process of the prior art.
Fig. 3 shows the relative energy efficiency/ reduction in
energy cost/reduction in greenhouse gas emissions for the
present invention compared to a steam extraction process
(SAGD) as a function of bitumen extraction temperature and heat
source temperature.
DETAILED DESCRIPTION OF THE PREFERRED
EMBODIMENTS
In this specification the following terms shall have the
following meanings. "Heavy oils" refer to crude oils, which have
high specific gravity and viscosity and are therefore difficult to
extract commercially because they do not readily flow. In this
sense heavy oil comprehends bitumen. "Waste heat" means
heat from a combustion process which is a byproduct of the
process. A "non-condensable" means any gas that is volatile at
a temperature and pressure at which the preferred solvent
condenses to a liquid. In this sense non-condensable means
any fraction that has a lower boiling point temperature than the
preferred solvent used in the extraction process. For example,
methane is a non-condensable gas in an extraction process
utilizing propane as a solvent. A "methane injection
specification" means the preferred amount of methane (or more
generically a non-condensable) present in a solvent which is
being re-injected and is preferably less than 5% and is most
preferably about 1 to 2 % or less. The preferred amount of
methane permitted depends upon the economics of purification
(removing methane from produced fluids) as compared to heavy
oil extraction rates (the negative effect such methane has on
CA 02374115 2002-03-01
7
production rates). A "heat pump" means any device or process
by which heat energy is transferred from some pre-existing heat
source or reservoir to the solvent to vaporize the solvent (and
typically is followed by a compression step to achieve desired
injection temperatures and pressures.). Thus heat pump heating
is contrasted with having to supply the heat by means of direct
energy consumption such as combustion or electrical energy
consumption or the like. A "latent heat source" means any free
or other heat source or reservoir that is existing but otherwise
dormant or undeveloped, including ambient air, geothermal,
ambient water or other sources and excludes any heat sources
involved in the direct consumption of energy such as
combustion heating, electrical resistance heating or the like. A
"secondary heat source" or "waste heat source" is any source
where heat is generated as a byproduct and otherwise would be
wasted if not used to vaporize solvent. "Heavies" means
hydrocarbon species with higher molecular weights than the
solvent. Thus if the solvent is propane, heavies comprehends
C4 and higher species such as butane, pentane and the like. A
solvent having a high vapour pressure at low temperatures is
any solvent that has a boiling point lower that about 20 degrees
C at atmospheric pressure. One preferred solvent is propane.
The "thermodynamic coefficient of performance" is defined as
the latent heat source temperature (in absolute terms) divided
by the temperature rise of the process.
Figure 1 shows a schematic process flow sheet for an
apparatus of the present invention. Figure 1 will be used as the
basis for a more detailed calculation of the energy requirements
of the invention. While Figure 1 shows a particular embodiment of
the invention, it will be understood that the present invention
comprehends such variations that may be necessary or desirable
to provide a greater or lesser energy efficiency at the expense of
capital cost, process stability and process control during upset
CA 02374115 2002-03-01
8
conditions. Figure 1 provides one preferred arrangement of
compressors, heat exchangers and pumps to provide a more
detailed estimate of process conditions and consequent process
energy requirements4. As explained in more detail below the
process steps include vaporization of the solvent at relatively low
pressure (temperature) and subsequent vapour compression,
prior to reservoir injection, which reduces the fuel requirements as
compared to a direct fired vaporizer.
By way of example consider a moderate size facility with
reasonable economies of scale has a bitumen production of
16,000m3/day (100,000bbl/day). The insitu gas/bitumen ratio is
assumed to be 2 m3/m3. The insitu gas is mostly methane and
typically has less than one mol percent of ethane. If a 55%
recovery rate of original bitumen in place (within an insitu
extraction chamber) is assumed, then the effective gas bitumen
ratio5 is 3.6 (=2/.55). Where the solvent is propane a preferred
reservoir extraction occurs at 50C. As will be understood the
solvent has a critical temperature that is higher than the formation
extraction temperature. The mean annual air temperature for the
Athabasca tar sands is about -6C, but it can be assumed that by
using heat storage in the summer and/or a variety of different
heat sources (geological etc), an effective annual average heat
source can be obtained at 5C. The produced solvent/oil ratio of
4.3 m3/m3 is determined by the heat balance at 50C, so
69,000m3 (of liquid equivalent) propane vapour is injected per
day. Approximately 3% propane makeup is required, the other
97% of the injected propane comes from recycled and
reconditioned solvent. Propane makeup includes about
1200m3/day is required for voidage replacement in the reservoir,
600m3/day of propane makeup is required due to propane
9 For the sake of clarity, the details with respect to
location and actuation of process control valves have not
been included in Figure 2, because isenthalpic valve
operation and actuation don't consume much energy.
5 In other words, 100% recovery of bitumen associated gas
but 55% recovery of the bitumen.
CA 02374115 2002-03-01
9
residuals in the sales bitumen and another 150m3/day of makeup
is required due to losses in the fuel gas6. Thus the total makeup
propane required is estimated to be 1950 m3/day.
Since the thermodynamic properties of propane determine
much of the energy requirements, the basis for the energy budget
in this specification is kilojoules of fuel energy per kilogram of
injected propane (kJ/kg). However the energy budget used can
easily be converted to a "per bbi of bitumen" basis through the
solvent/oil ratio. As will be understood by those skilled in the art,
the energy budget used is conservative in that it includes items
such as artificial lift for the produced heavy oil that might typically
be powered with electricity from the grid. In this manner, a more
thorough estimate of the true energy requirement (and the GHG
emissions) can be provided.
Figure 1 shows the produced fluids 10 flowing from the
production well(s) 12 into the free water knockout (FWKO) 14.
The FWKO includes a baffle 16 to separate most of the produced
water, which is disposed of through line 18. The well(s) 12 would
typically be operated in a control mode to produce liquid and
minimize production of solvent vapour.
The process parameters of the present invention depend
upon the starting conditions for the process. Thus, the present
invention comprehends that produced heavy oil may be either
artificially lifted to the surface, such as by means of a pump or
simply lifted to the surface by the pressure difference between the
well bore and surface pressure. The advantage of a pump is that
it is then possible to maintain the pressure in the FWKO to a
pressure slightly below the bubble point to minimize the vapour
overhead in the FWKO. As well the methane injection
specification needs to be maintained as explained in more detail
below. Altematively, if the pressure difference is used to lift the
6 Reference 2 predicts 20,000m3/day of propane is
available in Alberta to 2010 (i.e. excess of supply above
demand)
CA 02374115 2002-03-01
produced fluids, a portion of the production will be vaporized. The
latent heat of vaporization would chill the produced fluids and
thus reduce the FWKO temperature. Relying on the pressure
differential eliminates the energy costs associated with a pump as
5 well as permitting simpler well bore configurations but is prone to
slugging as the well bore will tend to randomly unload. The fuel
energy requirement for artificial lift is about 7kJ/kg propane7.
The produced fluids (heavy oil) include water, solvent,
bitumen and solution gas associated with the bitumen in the
10 reservoir. The FWKO drops out the produced water. Such a
separation is relatively easy compared to SAGD since the
produced fluid viscosity will be less than 100cP and the density
difference between the produced water and the produced fluids
will be quite large at 0.4g/cc (=1-.6).
Any gas which is present in the FWKO preferably flows to
a de-methanizer distillation column or the like. Since about 90
mol% of the hydrocarbon production is propane, the gas/liquid
ratio (and temperature) in the FWKO can be controlled by setting
the FWKO pressure. Since methane is the most volatile species,
it is preferentially concentrated in the overhead gas. Thus, by
setting the FWKO to a desired pressure the hydrocarbon liquid
stream will meet the methane injection specification while
simultaneously restricting the demethanizer size/duty to the
smaller proportion of the produced fluids in the overhead vapour.
For example, for a 1.3 mol% methane in the produced fluids, the
FWKO could be operated at 5 mol% vapour - 95mol% liquid.
This provides an overhead vapour with >6-mol% methane while
the FWKO bottoms would meet a 1-mol% methane injection
specification. To obtain a lower methane specification in the
liquids, then a lower pressure (temperature) can be used so that
a larger proportion of the produced methane is stripped out via
' This accounts for both the energy efficiency of the
driver/pump and lifting the bitumen as well as the
propane.
CA 02374115 2002-03-01
11
the FWKO overhead gas. For example, a 10 mol% vapour
overhead would reduce the residual methane in the FWKO liquid
to about 0.7mol%.
In the past it has been taught that extraction solvents
preferably include non-condensibles such as methane, ethane,
nitrogen or C02 in fairly large concentrations to achieve target
dewpoint specifications. In contrast, the present invention teaches
that the non-condensibles such as methane should be minimized
in the solvent since even small concentrations of less
condensable gases such as methane in the extraction solvent
dramatically reduce the energy efficiency of the separation and
purification of such solvents. However, methane removal comes
at a cost. For example, doubling the FWKO vapour overhead will
double the size and duty for the demethanizer and yet only
achieves a 30% reduction in the methane concentration in the
injection solvent.
The demethanizer overheads will typically contain
methane, ethane and a small amount of propane (<5%). The
overhead gas can for example be used for fuel. Since propane
losses in the demethanizer overhead represent an expensive
source of fuel gas (about twice the cost of methane per kJ of
heat), propane losses in this way are preferably minimized.
Absorber columns and the like may be used to increase the
propane recovery.
The demethanizer bottoms will provide near injection spec
propane. The demethanizer is a fairly energy intensive separation
due to the low temperature of the condenser (<-100C). Thus,
restricting the demethanizer feed rate via FWKO pressure control
will be helpful for controlling process energy requirements.
Furthermore, since the demethanizer condenser operates at
cryogenic temperatures, it will be necessary to dehydrate or
otherwise prepare the demethanizer feed to avoid ice (water,
C02 or hydrates) buildup. The demethanizer energy requirement
CA 02374115 2002-03-01
12
is estimated at -3MW, which is about 8kJ/kg of injected propane.
Referring to Figure 1 again, we see that the FWKO
hydrocarbon liquids flow into a series of propane evaporators.
Although three are preferred, more or less could be used as
desired. The evaporators use vapour recompression with
condensation (in other words an intemal heat pump arrangement)
to supply most of the latent heat of vaporization. This
arrangement with internal recycle (vapour recompression) helps
to reduce the energy requirement for separation of the propane
from the bitumen.
Referring to Figure 1 again, we see that the FWKO
hydrocarbon liquids flow into a series of three propane
evaporators. Each evaporator includes a fluid pump 50, a heat
exchanger 52 and a gas liquid separation vessel 54 and a
compresser 56. The gas liquid separator has a fairly large
diameter and may include internal features such as baffles and
demisters to facilitate gas liquid separation. The large diameter
of the separator 54 provides a fairly long residence time for the
fluids (i.e., slow superficial gas velocities) to help avoid liquid
carryover (i.e., mist or foam) into the gas overhead stream. The
overhead solvent vapour is compressed in the compressor 56
to an elevated pressure and temperature. The compressed
solvent vapour is then recondensed in the shell side of the heat
exchanger 52, releasing its latent heat of condensation to heat
and vapourize the bitumen solvent blend. If the temperature
difference across the heat exchanger 52 is kept to fairly small
values (i.e. 10 to 20C) then the compression horsepower
requirement is very small relative to the total latent heat required
to vapourize all the propane. For example with a 10C
temperature difference across the heat exchangers 52 we can
vapourize/recondense 125MW of propane for about 5 MW of
compressor power. The liquid draining from the bottom of the
separator vessel 54 is pumped though the tube side of the heat
CA 02374115 2002-03-01
13
exchanger 52, heated and flashed back into the separator
vessel 54. A portion of the liquid discharged from the pump 50
is also directed towards the next evaporator stage, while fresh
blend fluid is fed to the evaporator from the FWKO to offset both
overhead vapour loss + liquid discharge to the next stage.
This arrangement with vapour compression
recondensation effectively provides an internal heat pump. The
advantage of this internal arrangement is that the process
conditions can be set to optimum separation conditions without
being affected by availability of external heat sources and
potential seasonal variations. However, since the latent heat of
condensation of propane (per kg of propane) decreases as the
pressure (temperature) increases we must supply a small
additional amount of heat energy to close the heat balance for
each evaporator. The makeup heat requirement is minimized by
keeping the shell-tube temperature difference across the heat
exchangers small. Figure 1 shows the makeup heat energy
supplied from condensation of process propane vapor.
The latent heat of condensation of propane (per kg of
propane) decreases as the pressure (temperature) increases.
Thus, a supply of an additional amount of heat energy to close
the heat balance for each evaporator is required. Figure 2 shows
the makeup heat energy being supplied from the condensation of
process propane vapor.
As will be understood by those skilled in the art, it
becomes more difficult to extract propane from the bitumen as
the propane concentration in the bitumen decreases. This is
because the vapor pressure of the residual propane is reduced as
the propane concentration in the bitumen is reduced. Therefore,
the compression energy (work) requirement (per kg of propane
vaporized) increases as the propane residual becomes smaller.
Thus, according to the present invention it is preferred to
evaporate the propane in a multistage process, so that a
CA 02374115 2002-03-01
14
substantial portion of the propane can be vaporized at favorable
conditions (i.e. higher vapour pressure). Each successive stage
operates at lower pressure and/or higher temperature. It is also
desirable to leave some propane residual in the bitumen. It is
estimated that about 2 weight percent propane in the bitumen
would allow the bitumen to meet typical pipeline viscosity and
vapour pressure specifications without the need for additional
diluent. This is a useful result, since reference 2 indicates that
there will be insufficient diluent supply to meet pipeline bitumen
blend requirements after 2005. Of course the present invention
comprehends other residual weight percent of propane
depending upon the circumstances. In general the most
preferred weight percent is one which balances the cost of
recovery, and the end properties of the produced fluids.
The energy requirement for the evaporators is largely
determined by the compression energy requirement. The
compression energy requirement is determined by the condenser
heat transfer coefficients and exchanger surface area and the
temperature difference needed to supply the latent heat to
vaporize the propane in the bitumen. It is believed that an
economic optimum might require about 5 MW of compression to
provide the entire 125MW of latent heat required to strip the
propane from the bitumen for a process sized as outlined before.
This level of efficiency is achieved by using a relatively small
amount of compression (i.e. temperature difference -10C across
the propane evaporation heat exchangers). Accounting for
compressor efficiencies, the propane evaporators represent a fuel
requirement of about 30kJ/kg of injected propane.
It can now be appreciated what the benefits of the
evaporator configuration of the present invention are. The
particular arrangement of Figure 1 allows the propane separation
from the produced bitumen to occur at optimal process
temperatures, which are effectively detached from the heat
CA 02374115 2002-03-01
source temperature. For example, if a variable heat source (i.e.,
ambient air) is used to supply the latent heat for vaporization in
the evaporators, then seasonal temperature variations could
result in very low evaporator temperatures. This could lead to
5 operational problems with high fluid viscosities, low vapour
pressures high compression requirements, high gas velocities
and potential for foaming and/or mist carryover problems.
Obviously this concem would be mitigated if a warm and reliable
heat source (geothermal or the like) was available throughout the
10 winter months. Another aspect of the present invention is to
deal with the impact of residual non-condensable gases in the
FWKO liquids (i.e. the 1 mol% to 0.7 mol% methane discussed
above). The residual methane impurities raise the bubblepoint
pressure of the compressed vapour and thereby reduce heat
15 transfer efficiency. This means that additional compressor
horsepower and/or temperature drop is required to recondense
the evaporator vapour overheads.
There are several ways to help mitigate this problem as
comprehended by the present invention. One way is to provide a
continuous bleed/purge from the condensers (to the
demethanizer) to avoid accumulations of trapped methane.
Another way is to take advantage of the fact that since
methane is more volatile than the propane; the methane residual
in the liquids will decrease in each successive evaporator. Thus,
much of the residual methane in the FWKO liquids will be
concentrated in the vapour overheads from the first evaporator. If
a reliable heat source is available for the first evaporator, then all
of the overhead vapour from the first vaporizer could be
compressed directly to injection conditions, without
recondensation, thereby avoiding the difficulty of recondensing a
methane/propane mixture. Whether or not to recondense the
vapour from the first evaporator will depend on these
considerations but is also comprehended by the present
CA 02374115 2002-03-01
16
invention.
Figure 1 shows that the vapour overheads from the final
evaporator go through a partial condenser and the liquids then go
to a depropanizer column. The depropanizer provides a means
to recover heavies (C4+) from the solvent vapour and to use this
fraction as an additional diluent for the bitumen. While this
separation is shown for completeness, in most cases it is
expected that the C4+ fraction in the bitumen/solution gas may be
too small to be worth recovering. This C4+ concentration in the
reservoir gas may vary depending on the particular geographical
location. For the purposes of the energy budget calculations
contained herein, it is assumed that this C4+ separation is
unnecessary. Thus, the propane evaporators produce two
product streams, 1) liquid propane of sufficient purity to meet
injection specifications goes to storage and/or the propane
vaporizer and 2) produced heavy oil which goes to sales. As will
now be understood, the final energy input for the process is the
energy required to vaporize the propane liquid prior to reinjection
downhole as shown in Figure 2.
Referring back to Figure 1, we can see that two separate
parallel pathways 86 and 88 are provided to vaporize the
propane prior to reinjection into the reservoir. The major
proportion of the propane is vaporized via the heat pump 86,
and a smaller proportion is vaporized via heat exchange 88 with
flue gas etc. The heat pump as described by path B of Figure 2,
consists of a two stage process whereby the solvent is first
vapourized by heat exchange 80 with a latent heat source 82
typically at a temperature below the desired extraction
temperature. The second stage of the heat pump process is
compression of the vapour via a compresser 84 to the desired
injection temperature and pressure.
The second vaporization pathway as described by path A
of Figure 2 is shown at the far right hand of figure 1 is a more
CA 02374115 2002-03-01
17
traditional vaporization via heat exchange. This second parallel
vaporization pathway is also included to take advantage of
waste heat generated within the process (i.e. from the 38%
efficient engine driver for the compressor) which cannot be
otherwise easily used by the heat pump. Such waste heat
would typically include hot flue gases and other waste engine
heat that would otherwise be rejected in a radiator. This second
vaporization path includes a pump 90 and a heat exchanger 92.
Figure 2 shows a Molliere diagram for propane. The
vertical axis is the pressure and the horizontal axis is the
enthalpy. The diagram shows propane storage conditions
(saturated liquid at 5C) and the injection conditions (saturated
vapour at 50C). The figure also shows a 5C isotherm and a 50C
isotherm. The thermodynamic data for propane was obtained
from the NIST website of the US government.
Figure 2 shows two altemate paths labeled A and B
whereby the propane is vaporized (i.e. taken from storage
(saturated liquid at 5C) to injection conditions (saturated vapour at
50C).
Path A represents a conventional direct heating or energy
consumption process. In this case the propane liquid is first
pumped from storage at .55 MPa into a high pressure vaporizer
at 1.7MPa. The line is almost vertical because it takes very little
energy (<2kJ/kg) to pump the liquid propane to an elevated
pressure. Heat addition initially raises the temperature of the
propane to the saturation temperature of 50C. Additional heat is
required to vaporize the propane. At this point, the propane
vapour is at desired temperature (50C) and pressure (1.7 MPa)
for injection into the reservoir. This path requires an energy input
of 409 kJ/kg (the distance along the horizontal axis of path A).
Accounting for the fact that a direct fired heater is only about 90-
95% efficient due to flue gas heat losses etc, the total fuel
requirement is -450 kJ/kg propane. As can now be understood
CA 02374115 2002-03-01
18
the inherent inefficiencies in converting any form of energy into
heat and transferring the same into the propane mean that the
direct heating method results in a thermal coefficient of
performance of less than one.
Path B represents the more energy efficient path of the
present invention that utilizes a heat pump for transferring heat
into the solvent from a latent heat source. In this case, the
propane is vaporized by heat exchange with a convenient (i.e.
"free") heat source. For example, the heat source might be
ambient air with a temperature somewhat above 5C. The air is
chilled to 5C and the propane is vaporized at a temperature of 5C
to a vapour pressure of 0.55 MPa (=80 psia). The propane
vapour is then compressed to the extraction chamber pressure of
1.7MPa (=250 psia). If the compression is adiabatic, then the
compression path is almost parallel to the phase envelope.
There is a slight amount of superheating of the vapour when it is
compressed so the endpoint doesn't lie exactly on the phase
envelope and the final temperature is 54.4C. Note that the only
energy input provided in path B is the work required for vapour
compression (i.e.52.7 kJ/kg). Assuming typical compressor
efficiencies of about 38% (by accounting for energy efficiency of
the driver engine as well as compressor) then the total required
energy input for path B is about 140kJ/kg of propane vapour.
Thus, the heat pump offers energy savings of 69% relative to a
direct energy consumption process. As set out in more detail
below the present invention offers a thermal coefficient of
performance of greater than one and most preferably greater
than two or more.
However, Path B offers additional efficiencies. The
inefficiency of the engine driving the compressor offers potential
waste heat recovery from the flue gases and the radiator as a
secondary heat source. It would be reasonable to recover about
50% of the initial fuel value from waste heat in the flue gas and
CA 02374115 2002-03-01
19
the radiator = 70kJ/kg. The superheat of 4.4C (i.e. 54.4-50C) also
offers potential recovery of 11 kJ/kg via a desuperheater. These
additional sources for heat recovery can be used to vaporize an
additional 0.2 kg of propane (=81/409). Thus, by taking
advantage of these additional efficiencies, the overall energy
requirement of Path B can be further reduced to 117 kJ/kg of
propane (=140/1.2). Thus, an enhanced heat pump process
according to the present invention utilizing these additional heat
recoveries requires only 117kJ/kg compared to the direct fired
process (450kJ/kg), for an overall energy reduction of 74%
relative to propane vaporization in a direct fired or energy
consumption heater.
Referring back to Figure 1, two separate pathways are
provided to vaporize the propane. The major proportion of the
propane is vaporized via the heat pump, and, optionally, a smaller
proportion is vaporized via one or more secondary heat sources
such as a heat exchange with flue gas etc.
Heat for the heat pump portion of the propane vaporization
is preferably supplied by heat exchange with ambient air,
although the present invention comprehends all available sources
of latent heat. In the case of ambient air, the heat exchangers
could become fouled with frost if the surface temperature (i.e.
propane pressure) for the exchanger drops too low. Strategies to
mitigate frost fouling could include occasional pressure cycling of
the propane in the exchangers to periodically defrost the
exchanger and/or using altemate heat sources when air
temperatures get to cold. For example, if there is a previously
extracted well chamber that has been depleted, it may be feasible
to circulate water/gas through the chamber during the cold winter
months to recover geothermal heat and then switch to heat
exchange with ambient air during the warm summer months.
Alternatively, it may be economic to use waste heat from co-gen
facilities and/or cycle water/gas through depleted SAGD
CA 02374115 2002-03-01
chambers for heat recovery. Altematively, geothermal heat could
be supplied by circulation of water through a convenient aquifer.
Altematively, waste heat could be recovered from a gas
compression station on a pipeline. In any event, it is expected
5 that the operator might utilize one or more of these several
different latent heat sources.
A detailed calculation of the expected fuel requirements of
the present invention can now be understood. The fuel
requirements include artificial lift (7kJ/kg), demethanizer duty
10 (8kJ/kg), propane evaporators (30kJ/kg), and propane vaporizer
(117kJ/kg) for a total of about 162kJ/kg. The total fuel energy
requirement is 5600GJ/day (= 69,000 m3/day x 500kg/m3 x
162kJ / 1000, 000GJ/kJ). For the medium sized facility mentioned
above, this would require 151,000m3 of methane fuel/day, with a
15 total C02 emission of 3kg/bbl bitumen production (as compared
to 70 to 125 kg/C02/bbI for SAGD or synthetic crude). It is worth
pointing out that up to about one third or more of the fuel gas
requirement could be provided by produced associated gas.
The calculations above do not fully account for smaller
20 items such as lighting, computers and instrumentation sensors,
etc. However, these are believed to be small in this context.
Thus the present invention provides a method to separate, purify
and vaporize the solvent that reduces the process energy
requirement/energy cost/CO2 emissions by about 95% compared
to the currently available steam technology.
It should be pointed out that the example presented above
assumes "pure" solvent is injected and does not consider the
impact of solvent purity. As, will be understood by those skilled in
the art, a 1 mole% methane contamination in the injection solvent
raises the extraction pressure by 150kPa (to achieve a 50C
bubble point) and/or reduces the extraction temperature by about
5C (at 1.7 MPa). Similarly methane residuals in the solvent will
have an adverse impact on heat transfer coefficients in the
CA 02374115 2002-03-01
21
propane evaporators and thereby increase the process
compression energy requirements. Then present invention
comprehends that the methane specification may need to be
below 1% to provide optimum economics.
Considering a carbon tax of $200/ton and a fuel gas cost
of $3/GJ and a propane cost of $160/m3, the present invention
has about 7$/bbI lower operating expenses than a steam process
(=94%x$3 + 94%x$200x70x(12/44)x(2.2/2000)). Furthermore,
based on Duerksen and Eloyan8, one could anticipate an
additional 10 degree API upgrade for extraction with heated
propane vapor for an additional value added of $4.00/bbl (10 x
$0.40/bbl API). An additional 2API increase could be realized
from the residual propane, for an additional value added of
$.80/bbl. Additional savings of $1 relative to SAGD will arise
through the elimination of the diluent requirement and a 50%
reduction in pipeline transportation costs.
An additional cost for the current process arises due to the
fact that solvent that must be purchased and used for voidage
replacement in the extraction chamber, and some solvent is lost
to sales bitumen and fuel gas. At the present time, the fraction of
the solvent vapour that can be eventually recovered via blowdown
at the end of the project is unknown. Furthermore the solvent
inventory costs (i.e. amount and amortization period) depends on
the extraction rate. However, an upper bound on this cost can be
obtained by assuming 0% solvent recovery from final blowdown.
In this case, the cost of the solvent is -$3/bbI bitumen. If we
achieve 50% solvent recovery, then the net propane solvent cost
is about $2/bbl.
The current 25% resource royalty charged by the
provincial government is only applied to the operating margin
after allowing the operator to fully recover all R&D, capital,
8 Duerksen, J.H., A. Eloyan, Evaluation of Solvent Based
In-Situ Processes for Upgrading and Recovery of Heavy Oil
and Bitumen, 6th Unitar Conference Proceedings, 1995, Vol
1, pg 359
CA 02374115 2002-03-01
22
operating and interest costs. Thus, the entire capital cost of a tar
sands extraction facility is effectively paid by "avoided" royalties.
In this financial context, the operating margin becomes the most
relevant criteria to compare different extraction processes. The
present invention offers operating margins $5-11/bbl higher than
a SAGD process. At a royalty rate of 25%, the present invention
would provide an additional $2-3/bbl of revenue for the Province
Alberta. This is equivalent to 2 to 3 billion$/year of additional
revenue at an anticipated production rate of 2.6 Million bbl/day.
The actual operating margins for SAGD projects are
unknown. Reported supply costs vary from $8 to $14 per bbl9.
The historical average netback price is estimated to average
$13.50 (calculated from National Energy Board price data 1988-
2001 using FOB costs and extrapolated light-heavy differentials).
Thus, the "best" SAGD operating margin is expected to be about
$5/bbl. Thus, the solvent process described in this patent
application offers operating margins 2-3 times (i.e., $11-16 vs. $5)
higher than the "best" SAGD with a simultaneous 95% reduction
in emissions.
A further advantage of the present invention can be seen
by examining the criterion of royalties paid per kg of emissions.
The solvent extraction process taught herein offers an advantage
40 to 60 times higher than SAGD. The present invention is both
more profitable and has a much smaller impact on the
environment. Significantly, the present invention reduces the
process fuel requirement so that the marginal cost of using an
expensive but clean burning fuel such as gas is almost
insignificant (-10-1 5cents/bbl).
Consider an altemative technology such as application,
2,332,685, which teaches surface deasphalting of the bitumen
with combustion of the asphaltene residue as fuel for steam
generation. In this case, the fuel operating expense of $3/bbl is
9 Canada's Oil Sands A Supply and Market Outlook to 2015,
National Energy Board Canada, pg 37
CA 02374115 2002-03-01
23
saved and the API upgrade adds about $2/bbl (5API x $0.40) and
some pipeline/diluent costs are avoided ($0.50/bbl) for an
operating margin of $10.50/bbi. However, the C02 emissions are
effectively doubled by burning pitch instead of natural gas, so the
operating margin per kg of emissions is increased by only about
5% ($10.50/($5x2) =1.05).
The discussion above assumed that a latent heat source is
available at 5C for the process. However, as noted above, the
actual heat source may include a variety of different sources at
different temperatures and may vary seasonally. Figure 3 shows
the energy efficiency of the invention relative to SAGD, as a
function of heat source temperature and extraction temperature.
Interestingly, the more detailed energy calculation, as
summarized in Figure 3, using a heat source at 5C and extraction
at 50C, actually predicts a better (i.e. 96%) savings compared to
the earlier estimate of 94-95% savings above. This difference is
probably due to additional process inefficiencies and/or adverse
fuel characteristics implicit in the NRCan SAGD C02 emission
factor of 70 kg/bbl.
Figure 3 shows that the energy requirement for a 70C
extraction temperature is approximately twice as high as a 50C
extraction. Thus, the first 42C rise (from an initial reservoir
temperature of 8C to 50C) "costs" as much as the next 20C rise
from 50C to 70C. This non-linear behavior reflects the fact that
heat pumps have very high coefficients of performance for a
small temperature rise, but heat pump advantage decreases
rapidly as the required temperature rise becomes larger.
Figure 3 shows that as the temperature of the heat source
increases, the process energy requirement decreases. This
reflects the fact that less compression horsepower is required if
the propane is vaporized at a higher temperature (pressure).
Figure 3 also shows that the energy efficiency does not
achieve 100%, even if the heat source is at the extraction
CA 02374115 2002-03-01
24
temperature (i.e. 50C heat source with a 50C extraction
temperature). This reflects the fact that there are other additional
process energy requirements beyond vaporization which include
for artificial lift, solvent separation, solvent purification, pumps and
an appreciable temperature driving force is required for heat
exchange to occur.
Figure 3 shows a range of extraction temperatures from
40C to 70C. The optimum extraction pressure (temperature) may
vary during the extraction cycle (high initially to quickly grow the
solvent chamber, low during the final stages of extraction to
recover the propane inventory in the chamber). The energy cost
of raising the extraction temperatures from 50 to 70C is fairly
small (15cents vs. 30cents per bbl), but the propane inventory
cost (assuming 0% recovery) increases by $1.30/bbl, and solvent
bitumen ratio and the size (cost) of the surface facilities would be
increased at higher operating temperatures. Thus, the strategy of
high extraction temperatures/pressures makes particularly good
sense when the chamber is small, so that high initial production
rates are achieved while keeping the total solvent inventory
(costs) relatively small.
It can now be appreciated that the present invention is
most economical if the cost per kg of solvent for compression and
heat exchange is less than the cost of a direct fired heater. Thus,
for solvents where the vaporization temperature (pressure) is
lower, the compression energy requirement is higher so there is a
practical limit to the usefulness of a heat pump. For example, a
heat pump would not be useful for a steam extraction process
(i.e. SAGD) because the compression energy requirement for a
220C temperature rise would greatly exceed the latent heat of
vaporization savings.
However, for a solvent extraction process operating at
50C, with absorption of heat from a latent heat source at 5C
according to the present invention (i.e. an arbitrary but
CA 02374115 2002-03-01
"reasonable" heat source temperature), the thermodynamic
coefficient of performance is 6.2 (=278K/(323K-278K)). This
means that according to the methods and apparatuses of the
present invention one joule of work in the compressor can deliver
5 6.2 joules of heat at 50C. In reality, compressors are inefficient,
so only about 38% of the fuel energy consumed by the driver
engine is actually available for compression. However, for a low
temperature extraction process (i.e. at 50C) substantial process
heat can also be recovered from both the flue gas and the
10 radiator (about 50% of the initial heating value of the fuel). Thus,
for 100 Joules of compressor fuel energy, it is expected that
about 38Jx6.2 +50J = 285J of energy can be delivered as
vaporized solvent at 50C. So a practical coefficient of
performance would be about 2.8. The present invention
15 comprehends a coefficient of performance of more than one and
most preferably more than 2 or more.
By comparison, a direct fired boiler is only expected to
deliver 90-95% of the heating value of the fuel as heat due to
inefficiencies such as flue gas heat losses and control of the fuel
20 air ratio, etc. Thus, the present invention is expected to provide a
68% (=1-(1/2.8))/.95 reduction in energy cost relative to a direct
fired heater for solvent vaporization. Furthermore, this invention
also provides a 68% reduction of GHG (green house gas)
emissions relative to solvent vaporization via a direct fired heater
25 (burning gas).
If one considers that the heated solvent extraction process
offers potential energy reductions of 81% (as noted above)
compared to SAGD, and that the present invention teaches a
further energy reduction of 68% (i.e. of the remaining 19%)
through the use of a heat pump and various heat recovery
techniques, it is demonstrated that the present invention offers a
potential energy reduction of about 94% relative to SAGD10.
10 The reduction depends on both the extraction temperature
and the heat source temperature
CA 02374115 2002-03-01
26
Additional energy efficiencies arise from desuperheating of
compressed solvent vapour, and the ability to utilize convenient
low temperature sources of process heat. Additional energy
requirements arise from the energy required for solvent recovery
from the produced fluids and the energy required for purification
of the solvent to meet injection specifications as well as lift
supplied to pump the produced fluids from the reservoir.
By way of further example, if the present invention was
used to deliver the entire 2.6 million bbis/day of bitumen
production by 2010, then the total emissions from bitumen
extraction would be about 3-5 megatons/year vs 73
megatons/year for the best current SAGD bitumen extraction
technology or 125megations/yr for synthetic crude.
To put these numbers in context, the elimination of all
automobiles in Canada would reduce emissions by only 53
megatons/year (Annex C, pg. C-26 of reference 1). Current
estimates of Canada's excess emissions above the Kyoto
Agreement are about 200 megatons/year. Thus, the elimination
of 70 -120 megatons/year of C02 emissions, as taught by this
invention, provides a means to achieve about 50% of the
reduction mandated by the Kyoto accord and simultaneously
increase the profit margins of production.
The present invention provides strong commercial
incentives for implementation, as the expected operating margins
(revenue, net of operating expenses) and profits are expected to
be 2-3 times higher than the best current SAGD technology.
Higher profit margins will also enable a larger percentage of the
tar sands to be extracted profitably. Industry, govemment and the
environment should all benefit from this technology.
While the foregoing description has concentrated on
preferred embodiments of the invention it will be understood that
various alterations and modifications are possible without
departing from the broad scope of the attached claims. Some of
CA 02374115 2002-03-01
27
these have been discussed above, and others will be apparent to
those skilled in the art. For example, although the method and
apparatus of the present invention is most useful in association
with heavy oil, they also have application in enhancing recovery of
less viscous oils.