Thermal Recovery of Bitumen at Wolf Lake

Abstract

Summary For more than 20 years, BP Resources Canada Ltd. has faced the technical and economic challenges of oil-sands exploitation in the development of the Cold Lake oil sands of east-central Alberta. In 1985, the WolfLake 1 project began commercial bitumen production (7,000 to 8,300 B/D [1100 to 1300 m3/d]), with cyclic steam stimulation used to exploit the Clearwater formation. Construction of a 12,600-B/D [2000-m3/d) expansion started in 1987. Also in 1987, production started from a 22-well satellite that exploits the Lower Grand Rapids formation, a channel sand overlying the Clearwater formation. To develop this and other potentially commercial technology, four experimental projects were conducted: two steamfloods, a cyclic steam and combustion pilot in the Clearwater formation, and a single-well test in the Lower Grand Rapids formation. Experimentation to develop combustion as a follow-up process to cyclic steam with air lasted from 1979 to 1984 and with oxygen from 1983 to 1988. One of the Wolf Lake 1 satellites will be converted to combustion to obtain further operating experience. While operating experience has been obtained, numerical models have also been developed and used to improve the cyclic-steam operating strategy. These models, along with the formation properties and project histories, are described. Also, principles of the injection and production operations used in the combustion pilot are covered, and some solutions are offered on how these procedures overcame or avoided problems typically associated with combustion operations. Introduction Thermal recovery methods for the production of bitumen from BP Canada's Wolf Lake lease have been under development for more than 20 years. This lease, formerly called Marguerite Lake, is part of the Cold Lake oil-sands deposit of east-central Alberta (Fig. 1). The lease covers an area of about 117 sq miles [300 km2] (Fig. 2) and contains over 3 billion bbl [475X 10–6 m3] of bitumen in place in the currently exploitable Clearwater and Lower Grand Rapids formations. This paper reviews the geological and reservoir properties of the target formations and the experimental projects and studies that led to the Wolf Lake cyclic-steam commercial project. This project, which now produces 8,300 B/D [1300 m3/d] of bitumen from the Clearwater formation (following an expansion of 1300 B/D [200 m3/d] in 1986). is briefly reviewed. A more detailed review of the reservoir studies conducted to optimize the project. including numerical and in-situ stress modeling and the field tests to support these studies, follows. Combustion was chosen as a follow-up process to cyclic steam because it had the potential to double oil recovery. The combustion process tested at the Phase A pilot, from air through to pure oxygen, is briefly evaluated. Principles of the injection and production operations at the pilot are covered, and some solutions offered on how these procedures overcame or avoided problems typically associated with combustion operations. Geology and Reservoir Properties The bitumen-bearing reservoir sands under the Wolf lake lease are part of the Mannville group of the early Cretaceous Age (Fig. 3). These sediments have been divided into four units: the McMurray (D Unit), the Clearwater (C Unit). the Lower Grand Rapids (B Unit). and the Upper Grand Rapids (A Unit). The Clear-water and Lower Grand Rapids formations contain the major exploitable bitumen-bearing zones. while the Upper Grand Rapids contains gas-bearing zones. Clearwater Formation. The Clearwater is the most important single reservoir. It can be divided into three main units over most of the lease. labeled the C1, C2, and C3 sands, the geologies of which have been described by Visser et al. The total bitumen in place currently exploitable is 1.4 billion bbl [225X 10–6 m3]. The thickness ranges from 79 to 140 fl [24 to 43 m), while the net pay in the developed areas is about 75 ft [23 m]. In the developed areas. the C2 sand is generally separated from the C3 sand by a silty shale about 0 to 3 ft [0 to 1 m] thick, while the C1 and C2 sands are separated by 10 to 16 ft [3 to 5 m] of silty shale (Fig. 4). Both the Phase A pilot and Wolf Lake project produce from the Clearwater formation. Table 1 lists the average reservoir properties in these projects. Fig. 5 shows the temperature dependence of the viscosities of the Clearwater and other bitumens. Further Clear-water bitumen properties are reported by Mowell and Hambling. In the lease's northeast corner, the Clearwater reservoir is significantly different; it has a thicker net pay section, coarser grain, contains a gas cap over a portion of this resource, and has a lower-viscosity bitumen-40,000 cp at 60deg.F [40 Pads at 15deg.C] (Table 1). The bitumen in place is 480 million bbl [76:: 106 m3] under the gas cap and 370 million bbl [59X 10–6 m3) elsewhere. The Phase 1 and 2 pilots (discussed below) were developed in this area. Lower Grand Rapids Formation. The Lower Grand Rapids (B Unit), deposited as a barrier bar aligned northeast/southwest (NE/SW), ranges in depth from 1,165 ft [355 m] in the northeast to 1,245 It [380 m] in the south. It contains six main sands, four oil-bearing sands (B10, B8, B6, and B4) and two water sands (B14 and B12) (Fig. 6). The shales separating the sands are thin but provide effective barriers. The B10 sand is the main reservoir, having a net-pay thickness reaching 70 ft [21 m] in places. Porosities and oil saturations are generally higher than in the C Unit. Table 2 reports the average properties of the B Unit sands in the two tested areas. the single-well site and the Wolf Lake B Unit satellite. The bitumen is more dense and viscous than the Clear-water oil. The volume of bitumen in place in the B Unit is 1.2 billion bbl [190:: 106 m3), in proportions of 60, 37, and 3% among the B10, B8/B6, and B4 sands, respectively. Phase 1 and 2 Steamflood Pilots The Phase 1 steamflood pilot began as a two-well test (118 It [36 m) apart) with a temperature observation well 40 ft [12 m] from the injector. The test's initial objectives were to determine the rate at which steam could be injected into the Clearwater and to determine whether the oil could be effectively displaced and produced. Both the injector and producer suffered casing failures. Injecting 17,000 bbl [2700 m3) of steam caused breakthrough to occur at the producer in less than l month. Over a 2-month period, the test achieved a steam/oil ratio (SOR) of six. Phase 1 was then expanded to a five-spot, with an injector/producer interwell distance of 295 ft [90 m]. In this test, between June 1966 and Feb. 1968, 333,000 bbl [53 000 m3] of steam and 82,000 bbl [13 000 m3] of hot water were injected. The project achieved an SOR of 7.6. The test, however, was too small to permit much confidence in predicting commercial production performance and was complicated by two casing failures. Therefore, the qualitative information and operating experience were used to design the next project. SPERE P. 178^