Note: Descriptions are shown in the official language in which they were submitted.
206.~515
1 FIELD OF THE INVENTION
2 This invention relates to a process for treating a
3 bitumen/water emulsion to separate the emulsion components.
4 BACKGROUND OF THE INVENTION
The assignee of the present invention has recently
6 developed a novel process for recovering bitumen from
7 subterranean Athabasca oil sand. This process has involved:
8 - sinking vertical shafts through the overburden and
9 buried cil sand to penetrate into the competent
limestone underlying the oil sand;
11 - excavating horizontal drifts or tunnels from the
12 shafts in the limestone;
13 - drilling pairs of parallel wells upwardly from a
14 tunnel into the oil sand and then extending them
horizontally through the oil sand along its base,
16 the wells being closely spaced vertically (5-8m)
17 and being co-extensive;
18 - establishing fluid communication between the wells
19 by conduction heating of the span of formation
extending vertically between them, by circulating
21 steam simultaneously in the two wells to create
22 two closely spaced, parallel "hot fingers"; and
23 - then converting the upper well to steam injection
24 and the lower well to produced fluid production,
the steam being supplied from ground surface
26 through a line extending down through the shaft
27 to the upper well and the production being
2069515
1 produced from the lower well through a line
2 extending up the shaft to ground surface.
3 This steam injection/emulsion production process using
4 a pair of wells is referred to as a steam-assisted gravity
drainage (SAGD) process.
6 The product from the SAGD process is a hot emulsion,
7 under pressure, of bitumen and condensed steam or water,
8 containing some solids.
9 Hot emulsions of bitumen, water and solids are produced
in the conventional surface-mined oil sand operations of Syncrude
11 and Suncor. These emulsions typically have a temperature of
12 80C. The separation of bitumen and water in these plants is
13 carried out by diluting the emulsion with a bitumen solvent
14 (naphtha) and then centrifuging the mixture. The naphtha
dilution serves to increase the differential in density of the
16 water and bitumen, thereby enabling successful separation of the
17 emulsion components by centrifugation. However this is an
18 expensive approach as some of the naphtha is lost and the balance
19 has to be recovered - in addition, the centrifuges are expensive
to operate.
21 It would therefore be desirable to provide a simpler
22 process for separating the bitumen and water in the emulsion.
23 The bitumen product will likely be sold to nearby refineries,
24 such as those of Syncrude and Suncor; they specify that the water
content of the bitumen product should be less than about 5 wt.
26 ~-
27 It is the objective of the present invention to provide
28 such a process.
2069515 ~
SUMMARY OF THE INVENTION
2 In accordance with the invention, the hot emulsion
3 produced by the SAGD process is introduced into a gravity
4 separator vessel at a temperature greater than 205C, preferably
at about 215C, and at elevated pressure (typically about 3000
6 kPa). A chemical demulsifier of conventional type is also
7 introduced into the separator. A conventional reverse emulsion
8 breaker (REB) may also be introduced. The mixture is retained
9 in the separator vessel for sufficient time (typically an hour)
so as to enable substantially complete separation of the emulsion
11 components to occur. The bitumen product is drained through an
12 outlet in the base of the separator vessel and normally contains
13 less than about 5 wt. % water. The water is removed from the
14 separator vessel through an outlet higher than that of the
bitumen outlet. This water product contains little bitumen
16 (typically 2 wt. % or less).
17 The invention is based on the discovery that, with
18 increasing temperature, the densities of produced bitumen and
19 water remain about the same until about 125C at this point, an
increasing density differential becomes manifest. The density
21 of the bitumen at these +125C temperatures is greater than that
22 of the water. At a temperature greater than 205DC, preferably
23 about 215C, the density differential is sufficient to enable the
24 required degree of gravity separation to occur in a separator
vessel in a practical retention time. This end can be reached
26 without having to resort to the addition of solvent diluents,
27 although chemical demulsifier is required.
206~
1 Broadly stated, the invention is a method for
2 separating water and bitumen present in the emulsion fluid
3 produced from a steam assisted gravity drainage process being
4 applied to a subterranean oil sand reservoir, comprising:
retaining the fluid, in contact with an effective concentration
6 of demulsifier, in a separator vessel operated at elevated
7 pressure and a temperature of at least 205C for sufficient time
8 to effect gravity separation of the bitumen and water; and
9 recovering a bitumen production stream containing less than about
5% by weight water and a water production stream containing less
11 than about 2% by weight bitumen.
12 DESCRIPTION OE THE DRAWINGS
13 Figure 1 is a schematic perspective view of the present
14 assignee's underground test facility used to practice
bitumen/water emulsion recovery using the SAGD process;
16 Figure 2 is a plot comparing the density of SAGD water
17 and bitumen with increasing temperature;
18 Figure 3 is a schematic showing the high temperature
19 and pressure separator circuit used to test the invention;
Figure 4 is a side sectional view showing the separator
21 vessel of Figure 3;
22 Figure 5 is an end sectional view of the separator
23 vessel of Figure 4;
24 Figure 6 is an end view, at the line A, of the vertical
baffle at the inlet end of the separator vessel;
26 Figure 7 is an end view, at the line B, of a vertical
27 weir at the outlet end of the separator vessel;
206~515
1 Figure 8 is a plot showing flow rates into and out of
2 the separator vessel during the test;
3 Figure 9 is a plot showing separator vessel
4 temperatures during the test;
Figure 10 is a plot showing water cuts at the separator
6 vessel feed inlet, bitumen outlet and water outlet during the
7 test;
8 Figure 11 is a plot showing demulsifier and
9 clarifier/REB concentration in the separator vessel during the
test; and
11 Figure 12 is a plot showing bitumen cuts in the
12 production from the water outlet during the test
13 DESCRIPTION OF THE PREFERRED EMBODIMENT
14 The invention is supported by two distinct experimental
efforts.
16 In the first effort, the temperature dependence of
17 bitumen density was assessed. Figure 2 plots the changing
18 density of each of produced SAGD water and bitumen with
19 increasing temperature. As previously stated, the bitumen
density is about the same as that of water in the temperature
21 region 50 - 125C. However, at about 125C the density
22 differential between bitumen and water begins to increase
23 steadily, as shown.
24 The separation of SAGD water and bitumen was then
tested in the bench scale, flow through, high temperature and
26 pressure separator circuit shown in Figure 3. The separator
27 circuit was located in a tunnel in the underground test facility
28 of Figure 1.
206~
1 During testing, the wells 1 of the SAGD project were
2 in a "blowdown" stage. Therefore a positive dlsplacement booster
3 pump 2 was used to increase the emulsion pressure to about 3000
4 kPa. The pressurized emulsion was then routed through a gas
separator 3, to remove non-condensable gases, and heated to about
6 220C in a shell and tube heat exchanger 4. Chemical demulsifier
7 was introduced to the flow at one of the injection points shown
8 in Figure 3. The heated, pressurized emulsion was fed to a 610
9 mm diameter x 3048 mm long gravity separator vessel 5. The
incoming feed entered and accumulated in a feed chamber 6. It
11 overflowed through the port 7 of a baffle 8 into a main
12 coalescing and separation chamber 9. An interface weir 10 was
13 positioned at the far end of the vessel 5, directly ahead of a
14 bitumen outlet 11 and water outlet 12. A radio frequency type
interface probe 13 was positioned downstream of the interface
16 weir 10. When the probe 13 detected the bitumen-water interface,
17 a valve 14 on the water outlet line 15 would close and bitumen
18 would be flushed through the bitumen outlet 11. The emulsion
19 flow rate to the separator 5 and the bitumen flow rate from the
outlet 11 were measured by mass flow meters 16, 17.
21 Three inlet flow rates were tested, as shown in Figure
22 8. These were approximately 7.2 m3/d (total fluid), 9.6 m3/d and
23 17.5 m3/d. With a vessel capacity of 0.64 m3, the corresponding
24 residence times were 2.1 h, 1.6 h, and 9.9 h.
The nominal separator vessel design temperature and
26 pressure were 260C and 7000 kPa. The separator was maintained
27 at 2300 kPa until day 16. The pressure was then increased to
28 2850 kPa for the remainder of the test. The separator was
29 operated at approximately 195~C until day 18. The temperature
206~51~
1 was then increased in an effort to improve performance. When the
2 separator temperature reached about 205~C - 215C, significant
3 improvements in ~itumen and water cuts were observed. This
4 temperature range was maintained as the target for the remainder
of the test.
6 Figure 9 shows temperatures measured at various
7 locations in the separator vessel. Early in the test a
8 significant vertical temperature gradient was observed. On day
9 10 there was a 30C difference between the top and side
temperatures. Ten centimetres of fibreglass insulation were
11 added to the existing ten centimetres on day 12 and the
12 temperature difference was reduced to approximately 18C.
13 Increased insulation of exposed piping and fittings and increased
14 flow rates eventually reduced the difference to 10C by the
conclusion of the test.
16 ~he separator was operated with no internal coalescing
17 baffles until a shut down on day 32. At this time, a pair of
18 baffles were installed for evaluation. A test of an alternate
19 baffle design was carried out from day 71 until the test was
completed. It was concluded that the impact of the baffles was
21 minimal.
22 During the test, over 1000 samples were taken for the
23 detsrmination of bitumen and water cuts. A daily average of the
24 cuts is shown in ~igure 10. Bitumen and water cuts were
determined by drawing off a cooled sample and analyzed on ~ite
26 by standard oil field centrifuge methods. The samples were also
27 sent to a laboratory for Dean Stark analysis to confirm these
28 centrifuge results. Samples were obtained primarily from three
2~6951~
1 locations. The emulsion inlet, the bitumen outlet and the water
2 outlet.
3 The inlet emulsion varied between 55% and 75% by wt.
4 water. The fluctuation was due to normal changes in production.
The average was 68% water.
6 The water cut of the production from the bitumen outlet
7 11 varied between 5 - 15 wt. % during days 1-18, when the vessel
8 contents temperature was about 195C. Starting on day 18 the
9 temperature of the vessel contents was raised and reached about
215C on about day 20. After the temperature was so increased,
11 water cuts of less than 5% were normal at the bitumen outlet.
12 Obtaining "clean" water at the water outlet 12 was more
13 difficult. The average bitumen cut in the production from the
14 water outlet was high, up to about day 24. In this connection,
the separator vessel was initially operated for three days
16 without any chemical additives. During those days, the water cut
17 at the bitumen outlet was about 30%. The injection of Champion
18 x 8881 oil phase chemical demulsifier was initiated on day 4 at
19 injection point 1. Between days 4 and 18, concentrations ranging
20 from 200 ppm to 1200 ppm (based on total fluid) were tried.
21 Results were generally poor, as shown in Figure 10. Beginning
22 on day 18, the vessel temperature was increased and moving the
23 chemical injection point was tried. More specifically, on day
24 18 the inlet temperature was raised to about 220C. This change
was reflected in less than 5% water in the production from the
26 bitumen outlet; however the production from the water outlet
27 still had a bitumen content of about 20%. An examination of the
28 water outlet production showed that most of the bitumen droplets
29 1 Trade-Mark
; 9
206951~
1 were less than 2 microns in diameter. It was felt that these
2 fine droplets interfered with mixing of the demulsifier with the
3 bitumen. The fine emulsion may have been caused by shearing of
4 the fluid in a booster pump upstream of the separator vessel.
The original demulsifier injection point 1 was downstream of the
6 pump 2. On day 25, the demu]sifier injection point was moved
7 upstream of the pump to injection point 2. At the same time, the
8 demulsifier concentration was increased to 1700 ppm. Following
9 these changes, the bitumen cut in the water outlet production
decreased to about 5%. On day 26, injection of Champion ZB 153
11 water clarifier was commenced while continuing the demulsifier
12 injection. On days 30 and 31 the bitumen cuts in the water
13 outlet production decrsased to about 1% and the amount of water
14 in the bitumen outlet production was about 2.5%.
As shown in Figure 9, the vessel temperature fell from
16 215C on day 26 to 195C by day 30. The decline in temperature
17 began almost immediately following clarifier injection. It was
18 felt that the water soluble clarifier chemical was being
19 deposited on the heat exchanger tube surfaces. The process was
shut down during days 32-36 to steam clean the heat exchanger
21 tubes.
22 The operations to this point could be described as a
23 learning phase. The complexities and interplay between chemical
24 rates, injection point locations and vessel temperature had
become apparent. The demulsifier concentration still had to be
26 optimized in hopes of reducing it to a level lower than 1700 ppm.
27 The effects of moving the clarifier injection point and
28 experimenting with other clarifiers and reverse emulsion breakers
29 (REB) also remained to be examined.
2~69~1~
1 After restart on day 36, the separator inlet
2 temperature was maintained between 210C and 220~C until the end
3 of the test. With high temperatures now maintained, demulsifier
4 optimization was attempted. Between days 36 and 46 the Champion
X888 concentration was varied between 200 ppm to 1500 ppm. On
6 day 46 the rate was adjusted to 450 ppm and this concentration
7 was maintained until day 81. This resulted in water in bitumen
8 production stream cuts of between 2.5% and 5%.
9 Although the bitumen outlet production water content
had improved significantly, there was still as much as 10%
11 bitumen in the water outlet production stream. Inje~tion of a
12 new clarifier chemical (Champion ZXW108) was started on day 40.
13 The clarifier was added down stream of the heat exchanger
14 directly to the separator vessel at Injection Point 3. This
clarifier was tested until day 53, with poor results. Cuts in
16 the water outlet production were between 5% and 10%. The
17 chemical was replaced with Baker Oil Treating's R5302 reverse
18 emulsion breaker (REB) on day 53. Nater quality improved to 1%
19 to 2% bitumen, but results were not as good as those obtained
earlier with Champion ZB153 water clarifier. In a final attempt
21 to improve the water quality the REB injection point was moved
22 to the vessel inlet at Injection Point 4. Results improved
23 dramatically, due to better mixing of the REB. Figure 12 shows
24 a plot of bitumen concentration in the production at the water
outlet hetween day 58 and the end of the test (day 83). Cuts as
26 low as 275 ppm were obtained.
2~6~15
1 The defined treatment objectives of less than 5% water
2 in the bitumen production stream and 2000 ppm bitumen and solids
3 in the water production stream were met and exceeded at
4 temperatures as low as 205C, residence times of less than one
hour, and chemical injection rates of 200 ppm demulsifier and 200
6 ppm REB.
7 The main source of emulsion in the bench scale test was
8 created by shearing either in the booster pump or bypass control
9 valve. Addition of demulsifiers and reverse emulsion breakers
upstream of the point where the emulsion was formed was effective
11 in offsetting the formation of stable emulsions.
12 It was important to identify factors contributing to
13 emulsion stability, such as the droplet size distribution and the
14 presence of clay fines, and other interfacial constituents. The
presence of a clearly defined water/bitumen interface throughout
16 the test indicated that the process successfully overcame these
17 stabilizing factors. The combination of high temperature and
18 chemical treatment was effective in destabilizing all emulsions
19 encountered.
Predetermination of oil phase viscosity and density
21 relative to the water phase over the operating temperature range
22 was critical in understanding the conditions which affected
23 normal separation. The bench scale test and laboratory work
24 clearly defined the effect of temperature on separation. With
increasing temperature both the viscosity and density difference
26 were working to promote separation.
12
2~6~
1 Coalescence of water droplets in the vessel did not
2 appear to contribute significantly to the separation process.
3 This was indicated by the fact that the process appeared to work
4 effectively without coalescing baffles. The size distribution
of water droplets in the bitumen product could explain why ths
6 baffles were not effective. These small particles may require
7 more shear energy to coalesce than could be supplied in the
8 vessel.
