This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 170920, “Visualization of CO2 EOR by Diffusion in Fractured Chalk,” by øyvind Eide, SPE, Martin A. Fernø, SPE, and Arne Graue, SPE, University of Bergen, prepared for the 2014 SPE Annual Technical Conference and Exhibition, Amsterdam, 27–29 October. The paper has not been peer reviewed. This work demonstrates that molecular diffusion may be a viable oil-recovery mechanism in fractured reservoirs during injection of carbon dioxide (CO2) for enhanced oil recovery (EOR). The oil-production rate from diffusion alone, however, depends heavily on the distribution of CO2 within the fracture network and fracture spacing. A numerical sensitivity analysis, using a validated numerical model that reproduced the experiments, showed that the rate of oil production during CO2 injection declined exponentially with increasing diffusion lengths from the CO2-filled fracture and the oil-filled matrix. Introduction Compared with other EOR methods based on gas injection, CO2 injection has many beneficial properties, among them that CO2 lowers the gas/oil interfacial tension and reduces oil viscosity and density, resulting in increased oil mobility and oil swelling. The main drawback with injecting gas is the high mobility, a factor especially true in fractured reservoirs. The success of a potential CO2 EOR project increases when key driving forces for oil displacement during a CO2 injection are identified. In a highly fractured reservoir, molecular diffusion could be an important driving mechanism; however, it would require high fracture density. Molecular diffusion is the mixing of fluids caused by random motion of molecules, and its calculation is described in detail in the complete paper. Diffusion is often neglected as a production mechanism during modeling of CO2 injection in oil reservoirs because it is computationally expensive to handle and is assumed to be of minor importance. While this might be true for most conventional reservoirs, it is not true for heavily fractured reservoirs and in laboratory experiments, where the diffusion distances are much smaller. Experimental Setup Fluids. A one-component pure mineral oil (decane) was used as the oleic phase in the experiment to achieve first-contact- miscible conditions with CO2 at pressures below 10 MPa to eliminate the effect of oil composition and to promote repeatability between tests. Core Material. Outcrop Rørdal chalk from Denmark was used in this study as an analog for North Sea chalk reservoirs. It has a composition of mainly calcite (99%) and some quartz (1%) and consists mainly of coccolith deposits. The Rørdal chalk is a homogeneous chalk type with similar porosity and permeability values between sister core plugs. Core Preparation. A smooth-walled, artificial, longitudinal fracture was cut in Core Plug C1 using a band saw. A constant fracture aperture of 1 mm was maintained by use of a polyoxymethylene (POM) spacer. The spacer consisted of three large void compartments separated by support columns, and flow between each void was through grooves to allow for fluid transport. The permeability of the fracture was several orders of magnitude higher than that of the matrix. Experimental Setup. The scanner was a fourth-generation medical computed-tomography (CT) scanner with a scanning time of 3 minutes and a pixel resolution of 160×160 Μm and a slice thickness of 500 Μm. The core was first scanned to localize its position and to ensure a vertical fracture plane. The core was then vacuum evacuated before a 100% dry image was measured. A 100%- CO2 image was obtained after the core was fully saturated with CO2. The system was then depressurized, vacuum evacuated, and fully saturated with oil before the oil calibration scan was obtained. All 100%-saturation scans were performed at experimental pressure and temperature (10.7 MPa and 42°C, respectively). The mineral oil decane was used in all tests. The test conditions used ensured supercritical CO2 conditions and first-contact miscibility between CO2 and decane. The confinement pressure was kept 1 MPa above the line pressure.
Read full abstract