Search

Abstract The paper proposes an interpretation of the phenomena observed on several ARC (Accelerating Rate Calorimeter) tests performed with different crude oils. The ARC apparatus was used as a closed system and the gas formed at the end of the reactions were analyzed by chromatography. Low Temperature Oxidation (LTO) and Negative Temperature Coefficient (NTC) areas were identified during these tests. The proposed interpretation is based on an analogy with the reaction mechanisms occurring during light saturated hydrocarbon oxidation. The determination of kinetics parameters for oxidation reactions was investigated. In the meantime, limitations of the ARC device, when used as a closed system, clearly appeared. The conclusion is that not only the determination of basic kinetics parameters is generally not possible with ARC tests, but also that a serious selection of oil reservoir for air injection technique based on these tests seems questionable. Introduction Success of oil recovery by air injection greatly depends on knowledge of reaction mechanism between oxygen and oil in porous media. Oil oxidation during in-situ combustion involves numerous competing reactions occurring through different temperature ranges. Generally, in order to simplify the studies, investigators classify the reactions in three steps (Dubdub1 et al; Brigham2 et al). - First, the low temperature oxidation (L TO) reactions, which are heterogeneous (gas/liquid) and generally produce polar compounds. For some light oils, if the reservoir conditions are convenient, L TO can produce mainly carbon oxides and water at long time (Solignac3). - Then comes the medium temperature oxidation (MTO) : oxidation reactions are slow, therefore cracking/pyrolysis which leads to coke, becomes the main phenomenon. - Finally, high temperature oxidation (HTO) : oxygen reacts with unoxidized oil, polar compounds and coke (Belgrave4 et al; Solignac3) to give carbon oxides and water. This paper describes ARC experiments on crude oil oxidation and proposes an analysis of experimental data, using works (Ben-Ai'm5 et al; Le Chatelier6 et al; Newitt7 et al; Cathonnet8 et al) on light saturated hydrocarbons (CICIO). We compare heat rate behaviors measured by the calorimeter (Fig. 1) with zones of reactivity (cool and normal flames) occurring during oxidation of light saturated hydrocarbons in the gas phase.

Abstract Pump off controllers have earned a place in the technology of rod pumping. Many methods exist for sensing when a well has pumped off. Most widely used are techniques based on the surface dynamometer card or motor speed or production rate. This paper describes several methods for sensing pump off using the downhole pump dynamometer card. These include areas inside of the pump card, areas outside of the pump card, set point and liquid fillage, among others. Procedures for calibrating the controllers are described together with provisions for high fluid level recovery. Combining the liquid fillage method with variable frequency drives and eddy current drives is presented as a way of performing variable speed - no stop control. Pump card monitors (PCMs) hold promise of being useful devices which are easy to apply and comprehend. Introduction Application of pump off controllers (POCs) began about 25 years ago. These controllers have proven useful in saving power and minimizing wear and tear on rod pumping equipment. These savings are realized by stopping the unit when the pump is not filling satisfactorily. The POC has gradually evolved into more than a device to sense pump off and to stop the unit.1 Rod load based controllers can also sense abnormal rod loads and shut the unit down before the pump and/or rods are damaged. Controllers that monitor motor speed can also prevent damage to the prime mover and gearbox. Controllers that monitor liquid rate can also prevent damage to the surface stuffing box packing. POCs sometimes increase production slightly by detecting malfunctions sooner than ordinary surveillance techniques. At this writing, about 35000 wells worldwide have been equipped with POCs. The purpose of this paper is to share ideas concerning how the downhole pumpcard can be used to sense pump off. While creation of the downhole card is not simple, use of the pump card to sense fluid pound is. The complexity of computing the pump card can be hidden in software and hardware without troubling the user. The result is a pump off control concept that is more easily understood and applied. The downhole pump card is computed by solving the one dimensional wave equation Equation (1) Available In Full Paper subject to measured boundary conditions comprising time histories of surface rod load L(t) and surface rod position P(t). The resulting solutions are used to construct pump dyno cards of utility in well analysis and control. Oil producers began to be familiar with computed downhole pump dynamometer cards2 about 30 years ago. These cards have the advantage of simplicity. Only two types of cards have to be comprehended in calibrating a PCM. These are full liquid fillage and incomplete liquid fillage (fluid pound or gas interference). With anchored tubing, pump cards that indicate full liquid fillage (no fluid pound or gas interference) have roughly a rectangular shape (Figure I). At topand bottom of the stroke the pump is virtually stationary as load is transferred from rods to tubing and vice versa.

Abstract Horizontal drilling has revolutionized the petroleum industry's exploitation of low productivity reservoirs such as heavy oil, tight and naturally fractured reservoirs, reservoirs with gas and/or water coning problems, as well as EOR applications. This paper studied only applications of horizontal wells in low productivity conventional heavy oil reservoirs under primary production. Currently, there are many field applications of horizontal well technology in Saskatchewan heavy oil reservoirs. The performance from horizontal wells drilled in heavy oil reservoirs to date is generally very encouraging. Although horizontal drilling technology is generally understood. there are still some key questions concerning the effective application of horizontal wells in heavy oil reservoirs. The objective of this paper is to study the impact of key parameters of the horizontal well design using a complete reservoir model which incorporates pressure drop in the horizontal wellbore. Several practical cases were studied for heavy oil reservoirs with viscosity ranging from 100 to 10000 cP. The key parameters evaluated included wellbore length and diameter. The results will be described in terms of oil production, costs and economic profitability. Results indicated that longer horizontal wells were more desirable than shorter wells due to larger drainage areas translating to higher reserves productivity. Wellbore diameter has a more significant impact on reservoirs with heavier crude oil. The pressure drop in horizontal wells had no significant impact all productivity in heavy oil reservoirs. Therefore, it is more likely that ill the future, we will see longer length and smaller diameter horizontal well applications in heavy oil reservoirs. Introduction There are more than a thousand horizontal wells world-wide, and more than 150 horizontal wells are producing in Canada (1). Their popularity continues to grow. Applications of horizontal well technology in low productivity reservoirs, such as heavy oil pools have not only shown higher production rates compared to vertical wells (2), they also have tended to reduce sanding due to larger contact area with the formation causing a lower fluid velocity near the wellbore (2 and 3). In heavy oil reservoirs with water coning problems, horizontal wells have shown higher recovery than vertical wells (4). Although the success of the horizontal well technology in the heavy oil pool development has been very encouraging, there is still much discussion on how to design horizontal wells effectively for heavy oil applications.

Abstract The paper will present the field test results of two electromagnetic (EM) stimulation projects conducted in the L1oydl1linster heavy oil area. The two wells (AI-1l-48-25 W3M Lashbum and AS-6051-27 W3M Northminster) produce heavy crude from the Sparky formation. Detailed production, electrical and operational data was gathered and analyzed to quantify the effect EM stimulation had on each well's producing characteristics. 77Je main areas covered by the paper include production characteristics (primary and enhanced), inflow performance relationships (primary and enhanced), operational considerations and some of the related economics. The data discussed to this point has been restricted to reservoir heating. An additional review of a tubing heating electrical configuration test (ran on the Lashbum well) has been included. 17ze viscous nature of the crude oil being produced caused severe rod fall problems. Tubing heating was initiated to reduce the wellbore fluid viscosity and eliminate the rod fall problem. The paper will conclude with a general overview of the EM stimulation process as applied to the Canada Northwest Energy Limited locations and the potential for future applications. Introduction The possibility of using electrical energy to heat oil bearing formations has been attempted on several occasions over the past four decades. In general terms, the electrical (and electromagnetic energy) is converted to heat. Three basic types of heating are possible (dielectric, inductive and resistive). This discussion will deal with the resistive heating process. Canada Northwest Energy et al (CNW) contracted with EOR International (EOR) to field test the process in the Lloydminster area heavy oil formations. A majority of the heavy oil sands around Lloydminster are well suited to the application of electromagnetic (EM) heating. Many of the formations are relatively thin and not well suited for other secondary, tertiary and thermal recovery methods. The EM heating process as applied in the CNW pilot areas creates a heated zone in the near wellbore region. The application of heat significantly reduces the viscosity of the oil, improves the oil/water flow characteristics, and overcomes some forms of formation damage. The end result is an improved pressure profile in the near wellbore region and a corresponding increase in the oil production rate.

Abstract There are extensive deposits of heavy oil in Saskatchewan and Alberta which can only be recovered with low efficiency by conventional methods. Primary recovery in the best of these heavy oil reservoirs can yield about 6 percent of the OOIP. Waterflooding can improve the recovery, but only to an extent of f-2 percent because of the adverse mobility ratio. Thermal recovery processes using steam can be effective, but these are uneconomic in many Lloydminster type reservoirs because the deposits are relatively thin and heat losses to the overburden and underburden are excessive. One potential means for overcoming the problem of heat loss is to make use of vapourized hydrocarbon solvents at their dew point. This paper discusses the results of experiments with Tangleflags heavy oil and propane carried out in a large, scaled, physical model. It focuses on the area of potential problems encountered with the use of saturated hydrocarbon vapours as a solvent. In particular, it addresses the issue of precipitation of asphaltenes which may lead to reduced permeability and lower production rates. Rates, recoveries, gas to oil ratios and other important parameters will be presented and field performance of the process will be predicted from the data obtained in the scaled laboratory model. Introduction Production of heavy oils and bitumens from the extensive deposits in Alberta and Saskatchewan is hampered by the high viscosity of these materials caused by the high content of asphaltenes. After these heavy crudes have been produced, often by thermal methods, dilution with light hydrocarbons is required to facilitate transport through pipelines to upgrading plants and refineries. The cost of steam injection to mobilize the heavy crude is large, mostly because there are inherent heat losses associated with heating surrounding rock formations that do not bear oil. This situation is further exacerbated in thinner reservoirs. A new idea is to recover heavy oils using saturated hydrocarbon vapours under carefully controlled conditions which lead to separation of asphaltenes from the heavy crude. The asphaltenes are thus left behind deposited on the reservoir matrix while the much lighter oil is recovered. The handling, processing and transport of such in situ upgraded oil is easier and more economical. In our previous paper[l] we described a new process for recovering heavy oil with the aid of a saturated hydrocarbon vapour. In this paper we present further experimental data to support the feasibility of the process and explore the benefits of in situ upgrading by asphaltene removal and deposition on the rock matrix.

Abstract The present study consists of an experimental investigation of dispersion in inhomogeneous layered, isotropic porous media. Thanks to an in situ method of measuring electrical conductivity, we have studied two layered models. We examine four different layouts. The first sample has three layers of different permeabilities. The second has only two layers. Furthermore. two models with layers oriented perpendicular to the flow are considered. Besides normal "gaussian" dispersion, clear cut anomalies were observed. The experimental breakthrough curves are skewed as a consequence of tracer tailing. Tailed breakthrough curves are known from displacement in cores with local heterogeneity. Good agreement has been obtained between the experimental results and the semiphenomenological model of Coats Smith. Thus~ the mechanism involved in this anomalous dispersion is not trapping in the so-called dead end pores, but rather is due to different residence times because of permeability difference. Introduction The mixing of a tracer during flow through a porous medium involves the interplay between the two basic phenomena of dispersion, namely, molecular diffusion and mechanical (geometrical) dispersion. If two miscible fluids are in mutual contact. initially with a shaped interface, they will diffuse into one another because of random molecular motion according to the diffusion law of Fick1:: Equation (1) (Available in full paper) To account for the tortuous path of a porous medium, equation (I) was modified by Klinkenberg2, who used the analogy with electrical conductivity, thus: Equation (2) (Available in full paper) Do thereby corresponds to the bulk molecular diffusion coefficient for the fluid multiplied by the geometric factor (F Ø). The formation factor is F= σ o / σ, were σ denotes the electrical conductivity of a rock saturated with a fluid of conductivity σOØ. is the porosity of the rock. Molecular Diffusion is the dominant factor at extremely low or zero flow rate. At a higher flow rates, fluid flow plays a key role: there is a competition between the dispersive effect of the velocity gradients in the flow, and the various processes that homogenize the tracer (diffusion and transverse dispersion) across the porous media. The dimensionless Peelet number, which is the ratio of convective velocity U to the average velocity of molecular diffusion over a distance I characterizes this competition. Thus: Equation (3) (Available in full paper)

Abstract A series of experiments were performed in physical models saturated with heavy oils of 0.6 to 12.7 Pa.s. The oil was displaced by gravity drainage at constant pressure and it was produced using horizontal wells. The pressure was maintained either using a vacuum pump or by injection of an inert gas at low injection pressures. Computer Assisted Tomography was used for the visualization of the displacement front. Additional gravity drainage tests were performed in long vertical columns and in two dimensional glass micromodels. In this paper a detailed description of the performed experiments is presented. Production characteristics are shown and images of the gravity drainage process are a/so included. The recoveries varied from 20 % to 50 % recovery of the original oil in place. The pressure difference was maintained between 10 and 30 kPa. In the cases where sand parting was not observed, a uniform displacement front was formed. In the cases where parting was observed the experiment was temporarily shut in. Substantial amounts of oil were produced following the shut in period. Temporary shut in was also implemented at gas breakthrough with considerable increase in the oil recovery. Introduction In the Lloydminster area of west central Saskatchewan and east central Alberta, large reserves of heavy oil exist which can be mobilized at reservoir conditions. Conventional primary production of these reservoirs has utilized mainly vertical wells on relatively large spacings, resulting in recoveries of 5–10 % OOIP. Recent applications of horizontal wells in these reservoirs hold some promise of increasing recoveries, but thin pay intervals, low reservoir pressures and low GOR's limit the effectiveness of recovery mechanisms based on gravity, natural pressure depletion and solution gas drive. The work described in this paper considers the enhancement of gravity drainage in LIoydminster reservoirs through injection of immiscible gas and production via horizontal wells. In addition to presenting results of lab experiments in various physical models, an extensive review of relevant literature is included.

Abstract This paper examines the effect of horizontal wellbore hydraulics and flow direction in the recovery from idealized heterogeneous reservoirs subject to bottom-water drive. It is shown that wellbore hydraulics significantly influence the inflow along the wellbore. However, the total rate of the well is not affected by wellbore hydraulics. It is also shown that the heterogeneities give rise to different wellbore pressure drops for different flow directions. A novel multi-level technique where the coarse-grid. fine-grid and wellbore equations are solved separately was developed. Preliminary test runs with this technique show substantial reductions in CPU time when compared to the standard technique. Introduction The inclusion of wellbore hydraulics in reservoir simulation with horizontal wells has received much interest recently [1,2]. It was shown in [I] that wellbore hydraulics can have a significant effect on the inflow (flow into the well) along the wellbore. Although the prediction of the production from a horizontal well may be very similar with or without wellbore hydraulics. the inclusion of wellbore hydraulics yields a substantially different drainage pattern along the wellbore. To increase the resolution of the flow around the wellbore a hybrid grid system which consists of local elliptical refinement of the coarse Canesian grid was used. The studies in [I] were restricted to homogeneous reservoirs. This paper examines the effect of wellbore hydraulics and wellbore flow directions in the prediction of recovery from heterogeneous reservoirs. The reservoirs under study contain zones with permeability contrasts and shale streaks. This paper also describes a novel solution technique for converging the systems of coarse-grid equations, fine-grid equations and wellbore flow equations. In [I], all these equations were converged simultaneously using Newton's method. In this paper, a multi-level technique is described which allows the separate solution of the different systems of equations. Thus, the coarse grid equations, the fine grid equations and the wellbore equations were solved separately. Attempts have been made to solve separately the coarse and fine grid equations [3,4], but have not been very successful. The method described here is more comprehensive than those in [2] and [3] because the wellbore is also included in the calculations. The key to the success of this multi-level technique is the transfer of information between the different systems through judicious specifications of boundary conditions. The advantages of this technique are highlighted.

Abstract Use of slotted liner as a sand control device is widespread in SAGD operations in Western Canada. Operating temperatures in such thermal EOR wells can be extreme, sometimes exceeding 270 °C (518 °F), and the associated compressive axial mechanical strain imposed by constrained thermal expansion can load the pipe material beyond its proportional limit. Selection of an appropriate slotted liner configuration is critical to ensure structural stability (and hence sand control) is reliably maintained during operation. Mechanical properties of the tubular material at elevated temperature strongly influence the compressive stability of the liner structure, but lower-temperature properties also affect the ease of pipe slotting on a production scale, which is typically achieved by plunging thin saw-blades through the pipe wall. Common slotting issues include breakage or unacceptably high blade wear rates. This paper describes the development basis for a new tubular material formulation that is specifically optimized for thermal structural stability in SAGD applications without compromising slotting performance. Elevated-temperature mechanical properties are designed to prevent compressive buckling failures and to minimize strain localization potential. Results of analytical and experimental work (including stability analysis of the liner structure, thermo-mechanical material testing, and bench-scale slotting trials) are described. Introduction Ongoing rapid development of bitumen reserves in Western Canada has led to an increased focus on developing robust tubular design bases for extreme service conditions in in-situ recovery schemes such as SAGD. Specifically, cemented or frictionally constrained tubulars are subjected to thermallyinduced, deformation-controlled loading that leads to a unique set of design challenges and more stringent requirements for post-yield material response than those employed in traditional elastic design. Slotted liner is used as a sand control device in a majority of SAGD wells. Slotted liners used in thermal applications must provide a stable structure under extreme thermally-induced compressive loading in order to maintain wellbore access and to prevent excessive sand from entering the wellbore. While the deformation resistance of slotted liners depends on geometric attributes such as pipe wall thickness, slotting configuration, and slot geometry, material properties have a considerable impact on structural stability and localization resistance. Materials typically employed in slotted liner installations include API grades such as K-55 and L-80. However, the API mechanical property requirements for such materials are defined at room temperature, and the variability permitted by the API specification is large. Efforts to develop engineering design bases for liners in specific fields must consider the operating conditions of the application, and generally result in a more specific requirement for post-yield material properties than is offered by API. Hence, an API grade designation is not considered to be a sufficient description of material requirements for such tubulars. The manufacturing process for slotted liner involves plunging a series of thin sawblades through the wall of the tubular. Typical slot quantities are in the hundreds per metre of pipe length, and with individual well lengths of 600 m to 1000 m, an efficient slotting process is desirable.

Abstract An important step in the numerical simulation of in-situ combustion for enhanced oil recovery is the selection of the pseudocomponents used to represent the oil. Frequently the oil is separated according 10 distillation cuts. In addition, a hydrocarbon pseudocomponent called "coke" is used to present the heavy residue that results from the reactions immediately ahead of a combustion zone. While this approach provides for some of the general changes that occur during in-situ combustion. it provides a poor basis for representing the overall set of chemical reactions. Chemical principles and the results of several publications from diverse sources suggest that a modified SARA analytical approach presents a sound basis for representing in-situ combustion chemistry. The potential for such all approach is discussed and a set of pseudocomponents is proposed that could plausibly represent all the major physical and chemical changes that oil experiences in the in-situ combustion process. Finally. the steps that are foreseen to successfully implement such all approach are outlined. Introduction In-situ combustion may be a good option for enhanced oil recovery of many reservoirs. Unfortunately, its use is diminished by the lack of reliable methods for predicting performance. Usefully accurate prediction of performance almost certainly requires numerical simulation. The complicated interactions that control firefloods vary with oil and reservoir types and cannot normally be represented in simpler forms. Many of these interactions depend on the chemical reactions that dominate in-situ combustion. Consequently, the selection of a suitable set of pseudocomponents and reactions is vital to the establishment of an accurate numerical model for firefloods. In the search for a suitable basis to express the chemical reactions of fireflooding, a thought-provoking publication by Ciajolo and Barbella1was found. Their thermal gravimetric analysis data suggested that SARA (saturates, aromatics, resins and asphaltenes) analysis offered this sought-after basis for pseudocomponent selection. Various data in another publication by Bad2 provided a separate indication that the oxidation and cracking behaviour of a wide variety of oils corresponded to their SARA analysis. In this paper, the SARA-based approach is used to build a plausible and comprehensive reaction scheme that describes the chemistry of in-situ combustion.

Abstract Non-thermal enhanced oil recovery (EOR) techniques show good potential for recovering oils from the thin and shaly heavy oil reservoirs of Saskatchewan. Among the non-thermal processes, immiscible carbon dioxide injection holds the most promise of accessing these reservoirs. This technique, however, is much less developed than thermal methods. The process, if proven applicable to Saskatchewan reservoirs between three and seven metres thick, will access approximately 90% of the total oil-in-place. Considering the above, a multi-client experimental research program was initiated at the Saskatchewan Research Council. The objective of this research program is to evaluate various solvents in the context of the heavy oil resource and to investigate displacement mechanisms associated with immiscible gas injection processes. This paper is divided into two sections. The first deals with the characterization of a Kindersley area heavy oil The characterization includes analysis of stock-tank oil with and without additives, recombined reservoir fluid, and reservoir fluid plus carbon dioxide. The second section describes a scaled physical model and two displacement experiments conducted using a Lloydminster area heavy oil. The laboratory phase behavior data were generated to show the effect of pressure and temperature on carbon dioxide solubility, oil density and viscosity, compressibility, and swelling factors. The addition of 73.1 sm3/m3 of carbon dioxide at 7 MPa reduced the viscosity of the reservoir fluid at 25.5 °C from 819 mPa •s to 45 mPa •s, an eighteen-fold reduction. The same reduction in viscosity by thermal methods would require healing the sample to approximately 80 °C. The above oil under similar conditions, increased in density from 963.0 kg/m3 to 974.3 kg/m3 and swelled approximately 15%. Two scaled model experiments (secondary displacements) were conducted using a 10-cycle water-alternating-gas (WAG) process with a WAG ratio of 4: 1. In each run, the total mass of carbon dioxide injected was1.41 g •mole (0.53 PV at 2.5 MPa, 0.30 PV at 4. 1 MPa). These displacements indicated the immiscible carbon dioxide WAG process to be partially sensitive to the operating pressure in the range of study. More important is the relative volume of carbon dioxide, at experimental conditions, which dictates overall performance. Introduction Carbon dioxide flooding to be the primary non-thermal recovery process that hold promise of allowing access to the typically thin reservoirs in which most of Saskatchewan's heavy oil is found. Thermal methods are often inefficient and uneconomical because of excessive vertical heat losses, due to thin pay zones, and steam-scavenging by bottom water zones. Carbon dioxide may behave as a miscible or immiscible fluid when contacted with oil at reservoir conditions. For petroleum reservoirs, Holm(1) defines miscibility as that physical condition between two or more fluids that will permit them to mix in all proportions without the existence of an interface. If two or more fluid phases form after some amount of one fluid is added to others, the fluids are considered immiscible and an interfacial tension exists between the phases. Therefore, miscible displacement may not be applicable to all reservoir fluids.

Abstract Heavy oils and bitumens may be considered to be composed or chemical constitutive molecules belonging to two distinct categories, namely, the malthenes (soluble in 40 volumes of n-pentane) and the asphaltenes (soluble in roluene but insoluble in n-pentane). The spatial organization of malthenes and asphaltenes leads to the macrostructural and microstructural properties of which viscosity is of paramount importance for production, transportation and processing considerations., Such spatial organization causes an increase in viscosity with an increase in asphaltene concentration. Hence the high viscosity of Athabasca bitumen is, at least, partly due to its high asphaltene content (>15 wt.%). Experimental results suggest that viscosity depends not only on asphaltene concentration but also on the molecular weight of the asphaltenes. For example, higher molecular weight asphaltene units have greater tendencies to form aggregates of larger sizes and therefore, provide greater resistance to flow,. The process of aggregate formation is a complex one and is not well understood. However, this process can be reversed by breaking up the aggregates with the application of energy in the form of high shear or heat. We have examined the effect of ultrasonic energy on the viscosity or Athabasca bitumen and its mixtures witholher solvents and solutions via break up of asphaltene aggregates. In this paper, we report the results of our experimental work. Introduction Bitumens produced from Alberta tar sand formations such as The Athabasca formation have viscosities higher than 100,000 mPas (cP) at room temperature Pipeline specifications require viscosities of less than 120 mPas measured at 20 ºC. As a resulthey can not be transported through a heated pipeline and addition of lighter diluents to reduce viscosity prior to transportation via a conventional pipeline, the latter is widely used in Alberta. With the projected increase in bitumen and heavy oil production and the dwindling supply of light condensate, the diluent method of bitumen and heavy oil transportation will become increasingly difficult. Projections are that condensate supplies may not be sufficient enough to transport the growing volumes of Alberta bitumens and heavy oils by the early 1990's(1). As a result the search for other viable options has been underway for sometime. One notable development in this area is the oil-in-water emulsion technology, commercially known as the Transoil process, where the viscosity of the bitumen is reduced by forming an oil-in-water emulsion prior to transportation and subsequent recovery of the bitumen upon emulsion break up. The oil-in-water emulsion technology requires a two stage emulsification process at the upstream end and a de-emulsification process at the downstream end of the pipeline. The bitumen that is recovered from de-emulsification process essentially remains as viscous as the original bitumen and needs to be subjected to various upgrading stages. Successful field trials of this technology has been reported(2). One traditional viscosity reduction process involving low seventy thermal cracking, known as visbreaking, is superior to oil-in-water transportation to the sense that it not only reduces viscosity, but also upgrades the bitumen to some extent. This reduces the upgrading effort required in a downstream refinery.

Abstract The treatment of near-wellbore flow phenomena has a strong influence on reservoir simulation results for thermal processes. Using large well gridblocks can result in high predicted injection pressures, and inaccurate predictions of override and coning behaviours. Several new discretization methods have been proposed to efficiently provide more accurate modelling of the near-wellbore region. These methods, Cartesian hybrid grids, control-volume finite-element (CVFE) grids, and CVFE hybrid grids, have been implemented in a general purpose thermal reservoir simulator. In this paper, they are compared for the modelling of three thermal processes; single well cyclic steam stimulation, multi-well cyclic steam stimulation, and steam drive. Results show that Cartesian hybrid grids and CVFE hybrid grids provide an efficient means of accurately modelling near-well phenomena. In multi-well cyclic steam stimulation processes, the hybrid grids allow the use of cylindrical grids to accurately and efficiently mode/the near-wellbore region, while including the influence of adjacent wells with offset steaming. For steam drive processes, the hybrid grids predict earlier steam override and breakthrough times than a standard Cartesian or CVFE grid. The CVFE hybrid grid has the advantages of providing more geometric flexibility and a more consistent discretization at the boundary between the CVFE and cylindrical grid than the Cartesian hybrid grid. Introduction The treatment of near-wellbore flow phenomena has a strong influence on reservoir simulation results. In the current practice for field-scale simulation with communicating wells, a coarse Cartesian grid system is used to represent the reservoir, and wells are represented as sources or sinks within the coarse gridblock. This representation of wells is usually nor adequate, especially in thermal processes. For example, the ability to inject steam depends on the viscosity reduction of oil or bitumen by heat conduction and convection in the vicinity), of the wellbore. Thus, because the well is located within a coarse gridblock, an artificially high well injectivity has to be used to allow steam injection. Similarly, for producers, the coning behaviour is not properly handled in coarse gridblocks and empirical pseudo-functions, which are not always accurate, are normally used. For horizontal wells, the near-well phenomena have even a stronger effect on the simulation results because of the longer completion lengths. Accurate simulation of a field-scale thermal project with a Cartesian grid would be extremely costly due to the large number of gridblocks required to accurately model the phenomena occurring in the field. Small gridblocks are required near the well to model the rapidly changing properties (temperature, pressure, viscosity, saturation, etc.) in this region. To reduce the over-all cost or the simulation, gridblocks with larger volumes may be used away from the wells where changes occur less rapidly and have less impact on simulation results. Typical methods used to vary the volumes of Cartesian gridblocks result in large aspect ratios for many gridblocks which can lead to further discretization errors. For an accurate simulation of near-well phenomena, the best grid system is the cylindrical grid system because it follows the potential and stream lines near the well.

Abstract The use of a high density concentration of Bacteria immobilized in a packed column (the Packed Bed Slime Reactor) is described for the biodegradation of organic petroleum wastewater effluents. The column has been characterized in lams of liquid holdup, mass transfer to a flowing countercurrent air phase and liquid dispersion. Equations are presented representing an accurate mechanistic model of the biodegradation behaviour of the column for phenol, cresol and actual refinery wastewater. The column established a 200 micron thick slime layer of bacteria. Biodegradation of the phenolic pollutants was 100% efficient at a rate of 1.6*10−3 kg/m3 /s, significantly faster than any previously reported rates. Negligible air-stripping losses occurred during 100% biodegradation runs and the column responded swiftly to step changes m inlet concentrations reaching new steady states within three hours after a disturbance. Introduction Petroleum operations involve the handling of a great deal of water along with crude oil and/or petroleum products. Frequently this water phase undergoes intimate contact with the oil phase whether it be due to coproduction directly from the reservoir or due lo condensates from steam stripping columns within an oil refinery. Depending on the current value of crude oil, it may be economical to pump recovery wells with many times more water than oil while most oil refineries require a vast quantity of fresh water to process that same crude oil into useful products. A typical oil refinery will consume 18 barrels. of water for every barrel of crude oil it brings across its borders. Although much of this water may either be unspoiled or used as a secondary recovery fluid at production sites, frequently petroleum operations generate considerable quantities of polluted waters which must be safely discharged. The primary ingredients of discharge water and typical effluent concentration guidelines are listed in Table 1 (from Beychok1). Of particular concern for petroleum operations are the organic phenols which contribute a major portion of the dissolved hydrocarbons in the wastewater. This is because the aromatic ring proves to be simultaneously a very stable and toxic molecule making phenols recalcitrant to natural degradation and poisonous to most forms of life. This paper will focus on a novel design for secondary treatment of petroleum wastewaters. The packed bed slime reactor (PBSR) will be shown to be much more efficient than the aerated lagoon for biodegradation of phenols. The enhanced efficiency combined with the traditional but simple design will result in less money for construction and operation. The phenol biodegradation characteristics of an actual refinery wastewater effluent will be shown to be identical La that of artificial mixtures used in research studies.

Abstract In a steam-assisted gravity drainage process, two horizontal wells, superposed 16.4 ft (5 m) apart, are placed near the bottom of the formation. Steam is generally circulated through the wells to establish hot communication paths between them and to initiate gravity drainage. However, this steam circulation phase delays oil production. Resu.lts of a research program to investigate two novel methods to eliminate steam circulation and enhance the recovery of bitumen or heavy oil through horizontal wells are described in this paper. In the first procedure, paired horizontal wells were linked transversely with vertical channels to improve liquid drainage, and in the second procedure, naphtha was used as a steam additive to accelerate oil drainage and to increase ultimate oil recovery. The physical models used in this study were designed according to scaling techniques which insured that gravity drainage was the major mechanism of recovery. The scaling technique also ensured that field bitumen could be used in the model. Geometric similarity was maintained, and time was appropriately scaled. Six experiments were conducted which indicated that more than 50% of the initial bitumen in place could be recovered in less than/our years. Using vertical channels to enhance the drainage of fluids from the injection well t the production well improved oil recovery from paired horizontal wells. When one vertical channel was used for a 65.6ft (20 m) portion of the horizontal wells, the occurrence of a maximum oil-steam ratio was observed 180 d before that of a baseline experiment and the final recovery increased from 52% to 64%. When two channels were used for a 65.6ft (20 m) portion of the horizontal wells, the final recovery was lower than that for one channel (56% versus 64%) while the same early production was achieved. Hot fluids bypassed tire region between the two horizontal wells, reducing recovery from that region. The presence of naphtha in the steam further improved the rate of oil production and final recovery. When 5 wt% (0.05 kg/kg) naphtha in steam was used, thermal recovery was 82% compared with 64% for a baseline experiment using steam-only injection with one vertical channel. When the naphtha concentration in steam was increased to 10 wt% (0.10 kg/kg), the final recovery was lower than that/or the (0.10 kg/kg), the final recovery was lower thon thai for the test with 5 wt% (0.05 kg/kg) naphiha (71% versus 82%) A high. non-condensable gas saturation in the steam chamber blanketed the heat transfer interface between the steam and bitumen, reducing heat transfer across the interface. Introduction The application of horizontal wells in thermal recovery of heavy oil and bitumen has gained considerable interest in recent years as a result of advances made in horizontal well drilling technology and more efficient implementation of the gravity drainage process. In addition. horizontal wells have greater contact with the reservoir formation than vertical wells and, therefore, appear to be more economical.

Abstract Steamflooding with horizontal wells promises to be a highly effective method for recovering heavy oils from shallow, thin formations. As is often the case for poorly consolidated heavy oil formations, sand production could constitute a serious production engineering problem. For horizontal wells, this problem could be exacerbated as the sands/formation fines could settle inside, and choke off sections of a horizontal well completely - leading to loss of production. To alleviate this problem, it is often necessary to case the horizontal wells. This paper reports results of a series of scaled model steam injection experiments investigating the role of a perforated casing on the recovery performance. Fluid production rate, well pressure drop, cumulative oil recovery and cumulative oil-steam ratio for open and cased horizontal wells are determined and compared to determine the impact of the casing. It is concluded that a casing, though necessary to control sand production into a horizontal well, could lead 10 lower cumulative oil recovery and lower thermal efficiency for the recovery process. For example. oil recovery and the cumulative oil-steam ratio in the absence of a casing were 49.7% of IOIP and 0.62, respectively. The corresponding values in the presence of a casing were 42.9% and 055, respectively. Introduction With innovations in lateral and horizontal drilling technology, more horizontal wells are being used to recover oil and gas world-wide. It has been widely recognized that the main advantages of horizontal wells over conventional vertical wells include larger reservoir contact area, ability to reduce the effects of gas/water coning, as well as the higher probability to intersect systems of vertical fractures in fractured reservoirs. In thermal recovery applications, horizontal wells - due to their extended contact with the reservoir - promise high production rates and improved steam sweep efficiency. For these reasons, many thermal horizontal well projects have been started up, and field-tested in Alberta and Saskatchewan 1–3. The main production mechanism for the AOSTRA Underground Test Facility (UTF) in Fort McMurray, Alberta is steam-assisted gravity drainage. The pilot has been so successful that an expanded Phase B which involves more and longer horizontal wells is being implemented for commercial operation 1. For Esso Canada Resources Ltd. 's Cold Lake horizontal well project, the first phase of operation involves cyclic steam stimulation of vertical wells to provide steam injectivity. Steam is then injected down these vertical injectors to push oil towards the horizontal producer 2. Scepcre Resources Ltd.'s North Tangleflags, Saskatchewan pilot 3 also utilizes vertical steam injectors to mobilize and displace the heavy oil into a long horizontal producer. Sand production is perhaps the most widespread and costly production problem associated with many thermal recovery projects. The sand and formation fines could either impair production by choking the wellbore or reduce production by preventing the proper operation of downhole pumping equipment. Also, they will require the replacement of equipment because of erosion damage 4.

Abstract Saskatchewan's proven reserve of heavy oil is estimated at 4 billion m3. About 20% of this reserve is recoverable by enhanced recovery processes. Thermal processes such as steam flooding and in situ combustion are the techniques most frequently used for these relatively cold reservoirs that contain very viscous oils. Since most of the oil occurs in rather thin formations, severe heat losses to adjacent rock strata would invariably reduce the effectiveness of thermal processes. On the other hand, chemical processes such as alkaline waterflooding have proved quite successful in laboratory tests. A properly formulated alkaline flood is expected to be more efficient and cost effective. A variant of the conventional alkaline flood is being proposed for these reservoirs, particularly for those that contain crudes with API gravities above 15. In this modified scheme, a hot alkaline formulation is used to reduce the oil viscosity. At the same time, the viscosity of the alkaline solution is considerably increased by the addition of thickening agents. In this paper we present a methodology which can be used to screen several alkaline reagents to determine the type and concentration appropriate for a target reservoir. Dynamic interfacial tension data and comparative oil recovery data from linear displacement tests are discussed. Introduction The so-called heavy oil belt of Saskatchewan covers a wide area that stretches from Lloydminster in the north to Kindersley in the south. The estimated ultimate oil in place amounts to 4 billion m3. In a recent appraisal of this huge reserve, Reid[l] suggested that 20% of the oil in place could be recovered by using enhanced oil recovery (EOR) techniques. To date, only thermal processes which include in-situ combustion and steamflooding/stimulation are being applied to these reservoirs. However, since the density of the crude ranges from 11 to 20 °API and since the average temperature of these reservoirs is about 20 °C, the viscosity of the crude at reservoir conditions will be 1000 mPa.s or higher. Another significant characteristic of these deposits is the occurrence of the oil in zones of relatively thin sands. According to Selby et a1[2], the thinness of the formations coupled with the large depths and low permeabilities makes thermal processes unsuitable for about 50% of the heavy oil deposits of Canada, the U.S. and Venezuela. Among the various non-thermal processes, alkaline flooding appears to be the most attractive. Alkaline reagents are abundant and quite cheap compared with conventional surfactants. Their effectiveness for the recovery of acidic crudes has in fact been known since 1927 by virtue of the work of Nutting[3] and Atkinson [4] . It is now generally accepted[5-8] that alkaline reagents react with surface active materials present in the crude resulting in the in situ formation of surfactant soap species. The adsorption of these in situ generated surfactants at the oil/water/sand interface may result in a drastic reduction of the interfacial tension and/or a change in matrix wettability. The end result is the mobilization of residual oil trapped in the fine pores of the reservoir sand.

Fluids used in hydraulic fracturing are required to have several characteristics: high viscosity and low leak-off into the formation during the injection, good proppant carrying properties, rapid breakdown during the closure. and easy clean-up during the post stimulation phase, Fluids with complex chemical formulations has been developed to meet those specifications: linear and crosslinked polymeric fluids, energized fluids (foams), gelled acids .. Those fluids depart from newtonian descriptions, and have a history-dependent rheology (to take into account the effects of temperature and shear dependency, with eventually a time-dependency due to the cross-linking process). The rheology of the fracturing fluids is an essential component in the design of the stimulation treatments, and fluid flow equations dictate the lateral and vertical pressure distribution in the fracture. Algebraic techniques have been used to provide relations, between the fracture parameters (length, width, excess pressure or injection rate) and the fluid rheological description for power law fluids 1. Introduction This paper uses a numerical technique for the simulation of three-dimensional propagation of a fracture, and considers the effects of deviation from power-law on the final fracture c:haraderistics. Simpler models (P3DH2) have been developed in the literature, and the previous model provides a tool to validate them. We compare here the vertical growth rates predicted by the fully three-dimensional model with the ones used by a discrete P3DH formulation. Many fracturing fluids have been shown to deviate from the classical power law representation, and we examine the effects of using some of the more complex rheological models on fracture geometry and proppant placement in the fracture. The paper finally addresses the problem of the influence of fracture roughness on treatment parameters, and fracturing pressure decline analysis. FORMULATION AND EQUATIONS Assuming an isotropic elastic behaviour of the formations, And that the linear fracture mechanics assumption hold, a variational Formulation for the general three-dimensional fracture propagation Equations can be formulated3 by coupling the three sets of equations. Elasticity The formulation is based on a finite element formulation of a general boundary integral equation, and has the advantage wrt the classical finite element systems of reducing the dimension of the problem by one: Equations available in full paper.

Introduction Westcoast Transmission Company Limited has written ad installed a company wide, computer based, Maintenance Control System (MCS). Fort Nelson Gas Plant was used as a pilot and was the first location to completely install the system. The system was justified, designed and installed over a 3 year period between 1981 and 1984. This paper describes the system, how it was implemented and the benefits that were gained as a result.