Summary
A large number of heavy oil reservoirs in Canada and in other parts of the world are thin and marginal and thus unsuited for thermal recovery methods. Immiscible gas displacement appears to be a very promising enhanced oil recovery technique for these reservoirs. This paper discusses results of a laboratory investigation, including pressure/volume/temperature (PVT) studies and coreflood experiments, for assessing the suitability and effectiveness of three injection gases for heavy-oil recovery. The gases investigated were a flue gas (containing 15 mol % CO2 in N2), a produced gas (containing 15 mol?% CO2 in CH4), and pure CO2 . The test heavy-oil (14° API gravity) was collected from Senlac reservoir located in the Lloydminster area, Saskatchewan, Canada.
PVT studies indicated that the important mechanism for Senlac oil recovery by gas injection was mainly oil viscosity reduction. Pure CO2 appeared to be the best recovery agent, followed by the produced gas. The coreflood results confirmed these findings. Nevertheless, produced gas and flue gas could be sufficiently effective flooding agents. Comparable oil recoveries in flue gas or produced gas runs were believed to be a combined result of two competing mechanisms—a free-gas mechanism provided by N2 or CH4 and a solubilization mechanism provided by CO2. This latter predominates in CO2 floods.
Introduction
A sizable number of heavy-oil reservoirs in Canada1 and in other parts of the world are thin and shaly. Some of these reservoirs are also characterized by low-oil saturation, heterogeneity, low permeability, and bottom water.2,3 For example, about 55% of 1.7 billion m3 of proven heavy-oil resource in the Lloydminster and Kindersley region in Saskatchewan, Canada, is contained in less than 5 m (15 ft.) pay zone and nearly 97% is in less than 10 m (30 ft.) pay zone.4,5 Primary and secondary methods combined recover only about 7% of the proven initial oil in place (IOIP).1 Such reservoirs are not amenable to thermal recovery methods: heat is lost excessively to surroundings and steam is scavenged by bottomwater zones.6,7 The immiscible gas displacement appears to be a very promising enhanced oil recovery (EOR) process for these thin reservoirs. The immiscible gas EOR process has the potential to access more than 90% of the total IOIP.1,7 It could, according to previous studies,6–12 recover up to an additional 30% IOIP incremental over that recovered by initial waterflood for some moderately viscous oils.
For the development of a viable immiscible gas process applicable to moderately viscous heavy oils found in this sort of reservoirs, we selected three injection gases for study: CO2 reservoir-produced gas (RPG), and flue gas (FG) from power plant exhausts. Extensive literature is available on CO2 flooding for heavy-oil recovery, dealing with pressure/volume/temperature (PVT) behavior,3,6,7,13-15 oil recovery characteristics from linear and scaled models,3,6-8,10-12,15,16 numerical simulation, and field performance.17–19 However, only limited data are available on flue gas and produced gas flooding.20–22 To determine the most suitable gas for EOR application from laboratory investigations, we need knowledge of the physical and chemical interaction between gas, reservoir oil, and formation rock; and information on the recovery potential for various injection gases for a targeted oil. The test oil selected for this study was from the Senlac reservoir (14° API) located in northwest Saskatchewan (Lloydminster area). The PVT properties for the oil/injection gas mixtures were measured and compared. A comparative study of the oil recovery behavior for Senlac dead oil and Senlac reservoir fluid was carried out with different injection gases to assess their relative effectiveness for EOR.
Senlac Reservoir Geology
The Senlac oil pool is located within the lower Cretaceous sand/shale sequence of the Mannville Group. The Mannville thickens northward and lies unconformably on the Upper Devonian Carbonates of the Saskatchewan Group.
The trapping mechanism for the oil is mainly stratigraphic. The lower Lloydminster oil reservoir is a wavy, laminated, very fine- to fine-grained, well sorted, and generally unconsolidated sandstone. It exhibits uniform dark oil staining throughout, interrupted by a number of shale beds of 2 to 9 m (6 to 27 ft) thick, which are distributed over the entire reservoir. The reservoir is overlain by a shale/siltstone/sandstone sequence and lies on a 3 m (9 ft) thick coal seam.
The detailed reservoir (Senlac) data and operating characteristics are provided in Ref. 5. The reservoir temperature is 28°C (82.4°F) and the reservoir pressure varies between 2.5 and 4.1 MPa (363 and 595 psia). The virgin pressure of the reservoir at discovery was 5.4 MPa (783 psia) and the gas/oil ratio (GOR) was 16.2 sm3/m3 (89.8 sft3 /bbl). The reservoir matrix has a porosity of about 27.7% by volume and permeability of about 2.5 mD. The average water saturation is about 32% pore volume (PV). The pattern configuration for oil production is five-spot on a 16.2 ha (40 acre) drainage area. The estimated primary and secondary (solution gas and waterflood) recovery is 5.5% of the initial oil in place.
Experiment
Wellhead Dead Oil and Brine.
Senlac wellhead dead oil and formation brine (from Well 16-35-38-27 W3M) were supplied by Wascana Energy, Inc. The oil was cleaned for the experiments by removal of basic sediment and water (BS&W) through high-speed centrifugation. The chemical and physical properties of cleaned Senlac stock tank oil are shown in Table 1. The formation brine was vacuum filtered twice to remove iron contamination from the sample barrels.
Optimizing the development strategy of combined steam flooding & cyclic steam stimulation for enhanced heavy oil recovery through reservoir proxy modeling