Mobility Control Experience in the Joffre Viking Miscible CO2 Flood
- Derril J. Stephenson (Westcoast Petroleum Ltd.) | Andrew G. Graham (Westcoast Petroleum Ltd.) | Richard W. Luhning (Alberta Oil Sands Technology & Research Authority)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- August 1993
- Document Type
- Journal Paper
- 183 - 188
- 1993. Society of Petroleum Engineers
- 5.6.5 Tracers, 4.2.3 Materials and Corrosion, 5.2 Reservoir Fluid Dynamics, 5.1.1 Exploration, Development, Structural Geology, 7.4.5 Future of energy/oil and gas, 2.5.2 Fracturing Materials (Fluids, Proppant), 5.2.1 Phase Behavior and PVT Measurements, 4.1.2 Separation and Treating, 1.6 Drilling Operations, 5.8.5 Oil Sand, Oil Shale, Bitumen, 4.3.4 Scale, 5.4.9 Miscible Methods, 4.6 Natural Gas, 6.5.2 Water use, produced water discharge and disposal, 2.4.3 Sand/Solids Control, 4.1.5 Processing Equipment, 5.4 Enhanced Recovery, 5.4.2 Gas Injection Methods, 5.4.1 Waterflooding, 5.5 Reservoir Simulation, 5.3.2 Multiphase Flow, 5.5.8 History Matching, 3.2.3 Hydraulic Fracturing Design, Implementation and Optimisation, 5.6.4 Drillstem/Well Testing, 5.7.2 Recovery Factors
- 3 in the last 30 days
- 696 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 10.00|
|SPE Non-Member Price:||USD 30.00|
This paper discusses mobility control in the Joffre Viking field miscible CO2 flood. Since 1984, three injection strategies have been tried: water-alternating-CO2 (WACO2), continuous CO2, and simultaneous CO2 and water. The studies showed that simultaneous injection results in the best CO2 conformance. CO2-foam injection has also been investigated.
Vikor Resources Ltd. operates Canada's first miscible CO2 flood in the Joffre Viking pool northeast of Red Deer, Alta., Canada (Fig. 1). The reservoir is a Cretaceous, Lower Colorado group, Viking formation sandstone deposited as overlapping sand lenses.1 Table 1 shows the Joffre reservoir properties.
The field was discovered in July 1953 and developed with over 400 wells on 16-ha/well spacing. Primary recovery was by fluid expansion and solution-gas drive. During 1957, the field was unitized and a linedrive waterflood was begun. By the mid-1960's, 42% of the original oil in place (OOIP) had been recovered, and the field had reached its economic limit and was abandoned. In the early 1980's, laboratory tests and reservoir simulation studies determined that the Joffre Viking was suitable for miscible CO2 flooding.
The Joffre Viking Tertiary Oil Unit (NTOD) was formed in April 1983. In June 1983, the Alberta Oil Sands Technology & Research Authority (AOSTRA) agreed to fund 75 % of the cost to implement a miscible CO2 experimental pilot. Vikor is operator of the miscible CO2 project for AOSTRA and the working-interest owners. The pilot and subsequent experiments were successful, and commercial field development started in Oct. 1991.
Conditions that promote viscous fingering and gravity override were known to exist in the Viking formation before CO2 flooding. The CO2/oil and water mobility ratios at reservoir conditions are 22 and 23, respectively; CO2 density is ˜80% that of oil and 60% that of formation water; vertical permeability averages 0.05 µm2. Therefore, we expected that some form of CO2 mobility control would be required to enhance CO2 conformance in the reservoir. Water injection with CO2 is the most common method used to improve CO2 conformance; however, water injection can reduce oil recovery by shielding oil from CO2,2,3 The key to a successful flood is an optimum injection strategy that maximizes oil recovery with efficient use of available CO2.
During operation of the project, we experimented with different injection strategies to maximize oil recovery efficiency, including continuous CO2 injection, WACO2 injection, and simultaneous CO2/water injection. Laboratory and field experiments were conducted with surfactant to generate foams to control mobility.
WACO2 Pilot Operation
Fig. 1 shows the WACO2 pilot, which consists of two adjacent inverted five-spot patterns (Patterns A and B) drilled on 32-ha/well spacing.4,5 Development began in late 1982 with 1 year of production to establish baseline productivity before CO2 injection. Source-water injection increased reservoir pressure to more than minimum miscibility pressure (MMP) in the pilot area during the latter part of 1983. CO2 injection started in Pattern A in Jan. 1984; three 10% HCPV slugs of CO2 and water were injected alternately into Patterns A and B; while Pattern A was on water injection, Pattern B was on CO2 injection, and when Pattern A was switched to CO2 injection, Pattern B was switched to water injection. Because of the excellent quality of the Viking sand, injection of > 20 % HCPV /yr into the pilot was possible, permitting rapid pilot evaluation with commercial well spacing. In Sept. 1986, WACO2 was completed. Continuous water injection into Patterns A and B was started and continued until Aug. 1988, when a CO2-foam mobility test was initiated in Pattern B. Currently, the pilot is on waterflood and is nearing its economic limit.
The decision to use WACO2 injection in the pilot was based on results of a 1981 simulation study.6,7 That study suggested that a three-cycle, 1:1 CO2/water, 10% HCPV injection/cycle would be optimum for mobility control and for reduction of oil shielding in the presence of high water saturation. Advantages of the WACO2 process are that it offers a high degree of operational flexibility in the field, reduces the potential for corrosion, and reduces possible injectivity problems caused by simultaneous CO2/water injection.
Pilot performance during the WACO2 operation was analyzed through an extensive field monitoring program and reservoir simulation studies.8,9 The analysis gave insight into how injection fluids flow in the reservoir and how the WACO2 process controls CO2 mobility. Fundamental changes to our conceptual CO2 flow model were required to explain production behavior.
The pilot performance showed that the wells could be divided into two general categories on the basis of production response. Well 14-17 in Pattern B is representative of a Type 1 well. All pilot wells except one exhibited Type 1 behavior. Production of an oil bank that exhibits an initial spike of oil and then tails off is typical of Type 1 wells. Fig. 2 is a plot of the fractional-flow history of oil and CO2 for Well 14-17 and shows that the oil bank starts in Dec. 1984 and peaks at a 50% oil fractional flow in March 1985. Peak oil production rapidly decreases to 30 % fractional flow by June 1985, after which there is an almost linear decline in oil flow until the end of CO2 injection. No subsequent oil banks were observed during the second and third CO2 cycles. A moderate oil production increase occurred after the third CO2 injection cycle because of the sharp CO2 production decrease and subsequent increase in wellbore oil and water flow capacity. Fig. 2 also shows that some mobile oil was in the reservoir around Well 14-17 before the start of CO2 injection.
Type 1 wells also have a characteristic produced-CO2 profile; in Fig. 2, note that significant CO2 response occurs simultaneously with the oil bank arrival. Another Type 1 well characteristic is the CO2 response during the second and third CO2 injection cycles. In these cycles, CO2 production occurs almost immediately after the start of CO2 injection even though the injection well is 560 m away from the producing well and a 10% HCPV water slug is injected between CO2 injection cycles. CO2 production also drops dramatically after injection is switched from CO2 to water. Type 1 well CO2 response suggests that a separate flow path is established in the formation for CO2 and that water does not flush the C02/oil phase out of this flow path.
|File Size||1 MB||Number of Pages||6|