Abstract Cold production is a recovery process used in unconsolidated heavy oil reservoirs in Alberta and Saskatchewan, Canada. In this process, sand and oil are produced together under primary conditions. Oil production rates can typically increase by one order of magnitude when sand is produced. The production of sand into a perforation was modeled using a horizontal sand pack flooded with live oil. Previous sand production experiments were performed using dead oil. The pack was scanned with an X-ray computed tomography (CT) scanner. A wormhole (high porosity channel) developed within the sand pack starting at the production end as soon as the back pressure was decreased suddenly from 780 psi (5.38 MPa) to 500 psi (3.40 MPa). The wormhole was stable to collapse when the production pressure was decreased from 780 psi (5.38 MPa) to 500 psi (3.4 MPa) and maintained at that pressure for 3 hours. The wormhole developed within the high porosity region (lowest cohesive strength) of the sand pack indicating that worn-holes in the field will likely develop in the weakest sands which are normally the sands with little cementation and therefore more oil. Under this rapid depressurization, gas did not come out of solution while the back pressure was maintained at 500 psi (3.4 MPa). The wormhole collapsed when the production pressure was decreased to atmospheric pressure. This indicates that a sudden decrease in the bottom hole pressure in a well may lead to wormhole collapse. Introduction Cold production has been used with commercial success to recover heavy oil in the Lloydminster area of Alberta, Canada. High production rates have been reported for heavy oil fields under primary recovery 1to11 when large quantities of sand are produced with the oil. Several authors have attributed this high oil production rates to the formation of high permeability cavities 1,2,12,13 , channels (wormholes) 5to10 or both 11,12,13,16,17 . Solution gas drive is generally considered to be the main drive mechanism 1,2,3 . In many cases, solution gas drive by itself is not sufficient to explain the enhanced oil recovery. Significant increases in oil production occurred only when large quantities of sand were produced. Experiments in sand packs with live oil (without sand production) have shown that the permeability of the sand pack is not increased when gas bubbles are generated 3 . Tracer experiments between injection and production wells have been performed in the field by several operators 5to11 to measure the travel time between wells after considerable sand production has taken place. In general, this time was at least one order of magnitude shorter than that normally predicted for unaltered formations. These anomalously short times were explained by the formation of either fractures or wormholes. Theoretical models of cold production, which assume that a radial zones of dilated sand develops from a wellbore when sand is produced into a well, have been developed 14,15,16 . The assumption of a large dilated zone around a wellbore differs from our observation of a high permeability channel (wormhole). Materials The Clearwater sand used in the experiment was obtained from the collection tanks at Suncor's former Burnt Lake pilot project 4 .
Abstract Wormholes are believed to be generated during the initial phase of cold production and partly responsible for enhanced production rates when sand cuts are reduced to their minimum. In this work we present a theoretical model describing sand and fluids flow within a wormhole uniformly filled with sand. This model is based on the Darcy-Brinkman equations for fluids in mobile sand, a rheological equation for sand motion and conservation laws. Velocity profiles of sand and fluid are calculated and used to establish the relation between the pressure drop and the flow rates. An estimation of wormhole size is also provided. This model provides understanding of the basic flow behavior in a wormhole during the massive sand production stage of cold production. It also provides some key parameters for field scale numerical simulations of cold production. Introduction VISCOUS DRAG AND CRITICAL RADII The detailed mechanism of sand erosion is a complicated subject It involves the geomechanical properties of the sand, the thermal and rheological properties of fluids, and the nature of sand-fluid and fluid-fluid interactions. However complicated, the key to flux erosion is that it happens when the drag force overcomes the resistance of the sand matrix. The drag force should be mainly the viscous body force due to flow of fluids through the sand matrix (Darcy flow). The resistance of the sand matrix is mostly a surface force. It is due to cohesion, adhesion and contact forces. It holds a chunk in place against the body force. The thickness of this chunk, Δ, can be as small as the averaged grain size or orders of magnitude larger depending on the sand properties. In the former limiting case, "raining" of sand grains into a cavity occurs. In the latter biting case, micro-cracks in the sand matrix might be observed just before the erosion. Assuming fluid flow into a cavity of radius R , the above-stated criterion can be expressed as: Equation (1) (Available in full paper) where σ is the surface strength of the sand, and P the fluid pressure. For a spherical geometry, the pressure gradient is inversely proportional to the square of the radius if the fluid is incompressible. Therefore, we can integrate the right-hand side of the inequality (1). Denoting the reservoir pressure as P re , the pressure at the wall of the cavity as P , and the radius of the drainage zone beyond which the reservoir retains its original reservoir pressure as R s , we find: Equation (2) (Available in full paper) Based on sand production experiments [l-3] the chunk thickness is normally much smaller than the size of the cavity and the cavity size is smaller than the size of the drainage zone, i.e., Δ « R < R s . We therefore can expand the above inequality in a series of Δ / R s . Keeping the leading term of such an expansion, we obtain: Equation (3) (Available in full paper) The cavity will erode when inequality (3) is satisfied. The erosion will stop when the inequality is not satisfied. The equal sign determines a critical radius.
Abstract Cold Production is a recovery process used in un-cemented Heavy oil reservoirsin which sand and oil are produced together under primary conditions. Sandproduction is known to be necessary in order to better access heavy oilreservoirs. The production of sand into a casing perforation was modeledexperimentally using a horizontal sand pack. Heavy oil flowed through the sandand out the orifice at one end of the pack. The pack was scanned using an X-Ray CT scanner. A high porosity (53%) channel (wormhole) was observed to develop inthe sand pack above a critical pressure gradient. The sand CUI was 44% (byvolume) as the wormhole was developing. When the wormhole broke through theinlet, the sand cut decreased sharply. CT images taken at this time showed thatonly the loose sand within the wormhole started In be scoured away from the topdown. The experimental observations suggest that the high sand cuts (20% to40%) from wells at the start of cold production are due ID the growth ofwormholes while the sudden decrease in sand cuts (ID 1% – 3%) indicates thatthe wormholes stopped growing. The residual sand cuts observed in the field arelikely due to the scouring of the sand within the wormholes. INTRODUCTION Several heavy oil producers 1,2,3,4,5 have recognized that producingsand with oil significantly enhances primary recovery. The widespread use ofprogressive cavity pumps has been a key factor in making the Cold Productionprocess economical by reducing the damage to wells caused by the pumping ofsand. Elkins et al. 1 were among the first to suggest that highpermeability channels, which they called wormholes, develop in oil formationswhen sand is also produced. The wells were located in the Southeast Pauls Valley Fietd, Gavin County, Oklahoma. They inferred the presence of wormholesfrom tracer tests, caliper surveys of the well before and after sandproduction, permeability calculations and the large volume of sand produced. Vonde 2 reported a significant increase in oil production rates whenwells, initially completed with 16 mesh(0.40 mm) liner slots, were recompletedwith 250 mesh (6.35 mm) liner slots. The sand cuts also increased to 10% byvolume. The wells they investigated were located in the Cat-Canyon Field, California. Infectivity tests, performed by Amoco, Canada 6 , have shown that theconcentration of an aqueous solution of fluorescent dye did not change afterbeing produced from an adjacent well. In a separate laboratory experiment, these investigators observed that the dye was adsorbed completely after flowingthrough a sand pack. Since the concentration of the injected dye did not changesignificantly in the infectivity test they inferred that the dye did not flowthrough a porous medium. They concluded that the dye flowed through a channel. Undiluted slugs of dye travelled at speeds up to 7 meters/minute through whatthey believed were channel systems over 2 km in length that connected up to 12 wells 6 . The wells were located in the Elk Point/Lindbergh fields, Alberta, Canada. Communication between an injection well and a producing well 500 meters awaywas observed by Suncor' in their Burnt Lake field. Alberta, Canada.