Abstract Use of a horizontal well for steam injection, or for fluid production in thermal recovery, is of practical interest in field operations. Experiments were carried out in a high pressure steam model equipped with an axially placed horizontal well. Steam was either injected into the horizontal well, or into the surrounding porous medium, with fluid production through the well. A variety of experimental conditions were employed. Temperatures and pressures were measured inside the well and at various points in the porous medium. Several effects were clearly evident: gravity segregation of steam led to uneven heating of the porous medium. The distribution of heat along the well axis was quite uneven. Pressure drop inside the well was small, but it influenced fluid flow distribution in the surrounding porous medium. Oil recovery and pressure drop were plotted as functions of volume of steam injected. The production behaviour of the horizontal well could be correlated with temperature in the surrounding porous medium. Effects of well length, diameter, vertical position in the porous pack, oil viscosity, and pressure differential on oil recovery, oil-steam ratio, etc. were examined. Introduction Over 2000 horizontal wells have been drilled for primary production in Alberta and Saskatchewan. A smaller number is used for thermal applications. Such applications utilize horizontal wells for production, injection, or both. Examples are: Tangleflags and Cold Lake, where horizontal wells are used for production, with vertical injectors, several cylic steaming type operations where horizontal wells are used for injection and production, and SAGD (Steam Assisted Gravity Drainage) in UTF and other areas where horizontal wells are used for both injection and production. There are more applications where horizontal wells are used as producers than injectors. Flow of fluids and heat in the vicinity of a horizontal injector or producer is of interest in the above applications. The work described in this paper was concerned with an experimental investigation of fluid and heat transport around a horizontal producing well in a physical model that allowed for gravity effects. The flow of fluids inside the horizontal well was also studied via the measured pressure drop in the well. Several types of wells were tested. Effects of well length, diameter, and vertical location on performance were examined. PREVIOUS STUDIES The advantages of horizontal wells, viz. large contact area with the reservoir, low pressure drawdown, and possible mitigation of coning, are well-known, and discussed in many papers. Likewise, field experience has been also been described in numerous reports. The success rate of horizontal wells as producers is about 60% in Saskatchewan. The experience with horizontal injectors is much more limited. Flow hydraulics and pressure drop inside a horizontal well have been the subject of several recent studies. Plaxton 1 gives a comprehensive survey of treatments of pressure drop in horizontal wells, and reports an improved method. The only previous study pertaining to flow inside a horizontal well and in the porous medium in the proximity of the well, under steam injection, was reported by Doan 2 and Doan, Farouq Ali and George 3 .

Abstract Thermal recovery of heavy oil has been a tantalizing prospect for Saskatchewan's heavy oil reservoirs. Over the past 30 years, many thermal methods have been tested in Lloydminster area, ranging from electrical heating and cyclic steam stimulation (CSS) in the early sixties, to tailor-made steam injectionprocesses. Along the way, many in situ combustion tests were also carried out, with limited success. The question is: will thermal methods ever work in the thin, relatively deep formations of Saskatchewan, many now riddled with horizontal wells? The answer is yes, if one is willing to approach each specific field situation on its own merits, and design a thermal process to suit the conditions. Selected thermal recovery projects, specifically steam injection, in Saskatchewan are reviewed to determine what worked and what did not. For example, CSS did not have a chance in the thin formations, but it did work in thicker sands; in situ combustion was unsuccessful in many field tests, but it has been a qualified success in one or two cases; steamflooding as such has been unsuccessful, but it can be made to work under some apparently unfavourable conditions. The extensive laboratory work done supporting thermal processes is also mentioned. The paper does not give another screening guide. Rather, an attempt is made to identify the conditions for the application of steam in typical Saskatchewan heavy oil reservoirs. The conclusion is that there is definitely a place for thermal methods in Saskatchewan. Extensive horizontal well drilling has not made the task any easier, and good laboratory work is the key to success, as shown by current experience. Introduction Steam injection is the most successful heavy oil recovery method. At present, the total oil production by steam in North America is 530,000 BID 1 (84,200 m 3 /day), of which approximately 110,000 BID is in Canada. Most of this production is by steam. In situ combustion accounts for 2,520 BID in the U.S. and 5,200 BID in Canada. (The largest steam injection project is in Indonesia, producing 250,000 BID; in Venezuela production is 277,000 BID). But the application of steam injection technology is far from uniform in the different areas. In some cases, such as Peace River and Tangleflags, great ingenuity is used to make steam workable under adverse conditions. In other case, such as Cold Lake, the steam injection technology has been dictated by the reservoir response, and has evolved accordingly. This paper addresses the question of whether steam injection is likely to be viable in the thin formations of Lloydminster area, and if so, what forms it might take. In 1960's, cyclic steam stimulation (CSS) was tested in a number of thin sands in Saskatchewan. In most cases, the first cycle was relatively good, with an oilsteam ratio of over 0.5, but subsequent cycles produced very little oil, because solution gas, which provided the drive, was depleted. Desirable aspects of steam injection Steam is a unique fluid: it is the best heat carrier (with the exception of certain exotic substances), with a mix of sensible and latent heats.

Abstract Based on Pujol and Boberg's scaling criteria, a series of experiments on steam-CO 2 injection strategies was conducted in a high temperature, high pressure scaled model to evaluate oil recovery processes for bottom water reservoirs. The scaled model simulated 1/8th of a jive-spot pattern for a Cold Lake oil sand deposit of 12.8 m thickness underlaid by a 2.2 m bottom water zone. In addition to steam-CO 2 conrinuous injection. CO 2 soak prior to steam injection, hot water-CO 2 injection and CO 2 followed by steam injection (CO 2 -steam sequential injection) were evaluated. The results indicate that the co-injection of a gas with steam accelerates and improves oil recovery rates, as compared to steam-only injection. during the initial Slage of the process. The CO 2 -steam continuous injection resulted in a better performance (final recovery and oil-to-steam ratios) than that from steam alone or steam-CO 2 sequential injection. Soaking the reservoir with carbon dioxide prior to steam injection reduced steam injectivity due to blocking Of the bottom water zone with a high viscosity oil (due to asphaltene precipitation). Introduction The design of an effective oil recovery process for bottom water reservoirs depends on oil sands to bottom water thickness ratio, permeabilities, and oil saturations as well as on the injection/production strategies used, For favourable reservoir parameters, the bottom water zone can be beneficial in the recovery of high viscosity oil since it provides initial injectivity of hot fluids into the reservoir. Steam injection has been successful in bottom water reservoirs where the water zone volume is small relative to the oil sands volume 1 . Variations of the steam injection process, such as the creation of barriers at the oil-water contact 2 , injection of gas 3,4 and polymers 5 have been suggested in the literature. To thermally reduce the viscosity of Cold Lake bitumen to 200 mPas (200 cP), it is necessary to increase the temperature to 120 °C. A reduction in bitumen viscosity could also be accomplished by contacting the biturnen with suitable gases such as CO 2 , methane or propane, either as alone or in combination. For example, the reduction of biturnen viscosity to 200 mPas could be achieved at 85 °C by dissolving CO 2 in bitumen at 3 MPa The solubility of CO 2 in reservoir fluids also results in oil swelling, vaporization of oil and interfacial tension effects. All of these are beneficial in the recovery of bitumen from oil sands. Nasr et al 4,6 have published experimental and numerical results on the use of steam, steam-N 2 and steam-flue gas in the recovery of oil from Cold Lake oil reservoirs with bottom water zones, The results indicated that the processes are gravity dominated and steam injection now rate has a major impact on process performance, The sensitivity of the process to well completion interval, initial oil saturation in the bottom water zone. and bottom water-to-oil sand thickness ratio was published by Law and Nasr 7 . The present study provides an assessment of a series of experiments designed to evaluate CO 2 -steam injection strategies for improved oil recovery as compared to from steam-only.

Abstract Heavy oil reservoirs of Saskatchewan are characterized by thin pay, shaley sand, heterogeneity, low productivity and bottom water. Thermal recovery processes are not expected to be economically feasible for reservoirs containing approximately eighty percent of the heavy all in the province. The carbon dioxide immiscible flooding process 15 considered to be the most promising among the non-thermal recovery methods for these reservoirs. This paper presents a survey of the recent laboratory and field studies of the heavy oil-carbon dioxide immiscible process and the results obtained from phase behavior and core displacement experiments. Viscosity, density, bubble point pressure, Gas-oil ratio (GOR) and swelling factor for the reconstituted reservoir fluids and CO 2 -saturated wellhead oil and reservoir fluid were measured at 28 °C and pressures up to 8.27 MPa. A forty-fivefold reduction in Viscosity and a 16% Increase In the swelling factor were observed in the heavy oil-CO 2 system. Carbon dioxide once dissolved in the heavy oil at a high pressure remains in solution even when the pressure is reduced. Waterflooding of the oil-saturated core produced 31.5 to 46.2% initial oil-In-place (IOIP). The Incremental recovery from the CO 2 Injection Into the water flooded core was between 12.7 and 38.6% residual oil-in-place (ROIP) in the water-alternating gas (WAG) mode and between 23.1 and 33.2% ROIP for the slug process. The total recovery appeared to level off at 60% IOIP. The CO 2 injection into the core at the initial oil saturation recovered 29.9% IOIP in the WAG mode and 41.6% IOIP for a slug process where CO 2 was allowed to soak for 69 h. There appears to be considerable merit in soaking with CO 2 prior to water injection. Introduction Saskatchewan has an extensive gesture of heavy all. The proven resource 1.0 × 10 9 m 3 and the probable oil-in-place 3.4x10 9 m 3 . 1 The heavy all reservoirs of Saskatchewan ate characterized by thin pay, shaley sand, heterogeneity, low productivity and bottom water. Fifty-five percent of the Initial heavy oil-in-place is contained in reservoirs less than 3 m thick and ninety-five percent in less than 7 m thick. 2 Primary and secondary methods together recover less than ten percent of the initial oil-In-place. The development of appropriate recovery techniques specific to Saskatchewan reservoirs is necessary If the full potential of this resource is to be realized. Saskatchewan's heavy oil enhanced recovery projects have centered around steam injection and in-situ combustion. While steam Injection has been very successful, it is not expected to be economically feasible for those reservoirs which contain approximately eighty percent of the heavy oil. 3,4 In-situ combustion, although applicable to thin reservoirs, has been and marginally successful thus far. Recovery of heavy oil using carbon dioxide immiscible flooding is considered to be the most promising among the non-thermal EOR processes, particularly for those reservoirs having such undesirable geology as found in Saskatchewan. 4–6 Since 1982 the number of carbon immiscible projects in the United States has grown from zero to eighteen. 3