Temperature Distribution Ahead of an Advancing Steam Zone

It is difficult to produce oil from diatomite due to unique rock properties such as low permeability, high porosity, fine laminations, and fractures. Steam injection is a promising enhanced oil recovery method for diatomite, but the generation and injection of steam may lead to large energy cost and, potentially, well failures. Injection of hot water instead of steam provides a means to decouple pressure and temperature while still achieving the benefits of thermal recovery. This study investigates and compares oil recoveries from fractured diatomite cores (1.5 ″, 3.5 ″, and 5 ″diameter) using hot water and steam. Both hot water and steam injection at 200 °C were applied under spontaneous and forced imbibition conditions. Cores from two different reservoirs (A and B) were used. The B reservoir oil and brine formation were adopted for all cases. Synthetic steam boiler feed was used to create hot water and steam. An X-ray CT scanner was used to visualize in-situ phases saturations and characterize the porosity/oil saturation distribution. Wettability alteration toward water-wet conditions was observed clearly at elevated temperature (200 °C) for all cores. The Amott index, Iw, for hot water and steam injection around 200 °C with reservoir A conditions were 0.41 (A#1) and 0.36 (A#2), respectively. The final oil recovery from A cores at 45 °C and 200 °C were 62- 64 % and 78- 82 % OOIP, respectively. The A reservoir whole core with 5 ″ diameter at 45 °C showed greater oil recovery from spontaneous imbibition than the 1.5 ″ diameter core. The Iw and final oil recovery from core A with both steam and hot water injection at 200 °C was nearly identical. With B core, the Iw increased from 0.28 (45 °C, water) to 0.34 (200 °C, steam). The Iw of B core was equal to 0.33 after both hot water injection and steam injection. The final oil recoveries of both hot water and steam at 200 °C showed 76 % and 80 % OOIP, respectively. Both in situ visualization and effluent analysis shows that rock dissolution and fines migrations contribute the evolution of fracture and pore networks within the core. Therefore, results suggest that hot water provides the benefits of thermal recovery, in some regards, while decoupling pressure and temperature.

This study was condueted to qualitatively investigate the recovery of bitumen from Athabasea tar sand in a two-dimensional vertical model using a solvent (naphtha). Both homogeneous tar sand packs and tar sand packs with a high-permeability channel near the base, simulating a fracture connecting the inlet and the outlet, were considered in this study. The experimental runs were conducted utilizing naphtha injection, naphtha injection with conductive heating, steam injection, and naphtha injection followed by steam injection. It was found that naphtha injection is technically successful in recovering bitumen from a two-dimensional system. Recoveries as high as 73.5 per cent were obtained. The process, however, tends to be inefficient, because the richest sample recovered contained only 48 per cent bitumen. Recovery depends on the rate of displacement, being higher at lower rates. Gravity segregation is an important part of the recovery mechanism in homogeneous packs. Continued naphtha injection opens a flow channel, but sweeps the inlet portion of the pack more completely than the outlet region. Formation plugging may occur due to flocculation of the asphaltenes. The plugging was found to occur only after large volumes of injection, and thus it was not a major problem. Conductive heating at the base of the tar sand pack, combined with naphtha injection, increases bitumen recovery. However, when a highly permeable channel is present, conductive heating and gravity segregation are not very effective. As regards steam injection, the steam action is concentrated around the inlet, and does not directly contact the bulk of the pack. However, naphtha injection prior to steam injection in a homogeneous tar sand pack is highly effective in opening a steam-flow path. Thus, steam is concentrated more toward the outlet, and recovery is greatly increased. On the other hand, naphtha injection, prior to steam injection, when a highly permeable channel is present, is ineffective, because the naphtha is immediately vaporized and produced. The results of this study – being one in a series of tar sand projects – help point to the type of miscible-thermal process which is likely to be most successful in the in-situ recovery of bitumen from the Athabasca tar sands. INTRODUCTION THE TAR SANDS of Athabasca, Cold Lake, Peace River and Wabasca, located in northeastern Alberta, contain some 780 billion barrels of hydrocarbons"'. Although it is certain that, with an increase in oil prices, it will become economical to process tar sands by surface mining methods, it is equally certain that, for ratios of overburden to ore thickness greater than the present limiting value of 3.0, the vast majority of the tar sands will require the use of an in-situ recovery method. The present paper deals with such a method, involving the use of solvents in combination with heat. The results of an experimental project, one of a series on tar sands, are described and discussed. Background, General Considerations And Objectives It is well known that steam injection, if carefully engineered, can lead to economic recovery of heavy oils.

Numerical study on steam injection in a turbocompound diesel engine for waste heat recovery

MARTIN, JOHN C., MEMBER AIME, CHEVRON RESEARCH CO., LA HABRA, CALIF. Abstract A theoretical analysis of steam stimulation is presented for single sands. The analysis includes the eject of steam production and most of the effects of heat conduction. The results show the effects of a number of important variables on the performance of an idealized well under steam simulation. Calculated responses are presented which indicate the effects of steam production, amount of steam injected, water production, formation thickness and formation damage. Results indicate that steam production can cause large reductions in the heat contained in the formation. This effect can be eliminated by drawdown control. Water production reduces the amount of oil produced during stimulation. The optimum amount of steam to inject depends on economic factors as well as the well response. In many cases, the increased temperature resulting from stimulation reduces oil viscosity near the well sufficiently to overcome the effects of formation damage even if the damage is not removed during steam injection. Calculated responses for thin sands are more favorable than anticipated. Introduction Little has been published on the theory of steam stimulation despite the interest it has created and the wide variation in well responses. The results of the present analysis provide an insight into steam stimulation, and the methods employed provide a foundation for future work. Analyses presented in Refs. 1 and 2 are very limited and apply to gravity drainage conditions. Ref. 3 contains an analysis similar to the one presented here. The idealized models used and the assumptions made in Ref. 3 are different from those used in this paper. The analysis assumes that after steam injection has heated a small portion of the volume within the radius of drainage of a single uniform sand, a shut-in soaking period is allowed before returning the well to production. The effects of gravity, capillarity, transient pressure and water-sensitive sands are neglected. The injection and soaking times are assumed short compared to the stimulated production time. The initial temperature is assumed uniform; thus, the results apply primarily to first-cycle stimulation. The effects of gas production other than steam are neglected, and the water-oil ratio during production is assumed constant. Steam stimulation involves the simultaneous variation of the temperature, pressure and saturations. General mathematical equations for these variations are complicated and can be very difficult to solve. Simplified equations based on idealized models are used to reduce the mathematics sufficiently to allow approximate solutions to be obtained. Discussion INJECTION An idealized model for which heat conduction is neglected is used to represent the behavior of a well during steam injection. The mathematics for this model is presented in Appendix A which also contains an approximate solution for the behavior of the no-conduction model. SOAKING During soaking, the well is shut in. Only the temperature, pressure and saturation distributions at the end of soaking are needed in the analysis. During soaking the heat is considered to be conducted in a uniform medium from an initially uniformly heated circular cylinder confined to the producing interval. At the end of soaking the saturations and pressures are assumed to correspond to the cold zone. Analysis of the heat flow during soaking is included in the next section. The radius of the heated cylinder is calculated from the following heat balance (for constant quality steam injection). r = [ .........(1) At the end of the soaking period it is assumed that little or no free gas is present near the well, and that the soaking period has been sufficiently long that the steam zone has had time to expand and the steam has condensed. The condition where there is no soaking is considered in the next section. PRODUCTION In this section, an approximate method is presented for solving the equations of heat and fluid flow associated with the production of oil and water during steam stimulation. Where initial pressure drawdowns are sufficient to cause steam to flow into the wellbore, steam production is assumed to occur within a short initial adjustment period (Appendix B). The production practices followed soon after the well is returned to production can have a large influence on the amount of oil produced during the stimulation cycle. Under most conditions, there is a period of time in which some or all of the water in the heated zone is converted into steam and produced. JPT P. 411ˆ

The petroleum industry has long been interested in potential methods of increasing recovery from oil reservoirs. The energy necessary to produce a satisfactory increase in recovery has been supplied through use of a secondary fluid to displace, heat, and/or dilute the original reserve. This paper describes a laboratory investigation into the application of electrical energy as a recovery agent in impermeable tar sands. The preliminary study presented has shown that flow continuity can be established and maintained in laboratory tar sand models solely by application of electrical energy. Formation of an electrically carbonized zone of 10 darcies permeability, high oil recovery, moderate thermal-productive requirements, and the production of qualitatively important proportions of electrocarbonization gas have been demonstrated with linear laboratory tar sand models. The size of the permeable channel created was found to be a function of the amount of electrical energy applied. Principal mechanisms responsible for these results have been identified as electrolytic and electronic conduction, electrical resistance heating, viscosity reduction, thermal expansion, and distillation with pyrolysis. Theoretical considerations on the effect of electrolyte content on the over-all resistivity of rock and electrode-earth contact resistance and heating rates arc discussed. Preliminary results of electrogasification and dielectric breakdown experiments with tar sands are mentioned. Introduction Application of thermal energy, as a means of increasing oil recovery from petroleum reservoirs, has received considerable industry interest during the past few years. The problem of producing highly viscous oils and tars presents an even greater challenge because of their tremendous reserve and because thermal recovery by fluid injection methods is not applicable for situations where sufficient flow continuity between wells cannot be established or maintained. Thermal recovery techniques which have received greatest attention are the injection of air to support underground combustion, the injection of steam, the injection of-hot water, and detonation of thermonuclear devices within oil bearing formations. Use of electrical energy has recently been considered for bottom-hole heating within a single wellbore in order to initiate combustion, or to heat injected or produced fluids. Standard resistance heating elements lowered by conductor cable are usually selected for such purposes. An alternate technique is to utilize the natural earth resistance between two electrodes spaced a considerable distance apart. This technique has been successfully applied to electrocarbonization" experiments with coal and oil shale. However, no recent laboratory work has been reported in application of this technique to the recovery of crude oil from sands. In the coal and oil shale investigations, the electrocarbonization treatment was found capable of creating a tortuous and permeable fixed-carbon channel, or "electrofracture", through the otherwise impervious strata. Somewhat similar results have been reported by Bill and Davis for oil sands and shales. In their experiments however, dielectric breakdown was used to create the permeable-conductive channel. They also found potential requirements for dielectric breakdown of an oil bearing formation was dependent on electrode spacing which precludes commercial application. On the other hand, electro-linkage - electrocarbonization processes have demonstrated their feasibility by successfully linking through 150 ft of coal with a maximum applied potential of 2,500 v. Total energy requirements for specific electro-thermal applications have been estimated for sands and commercially measured for oil shale.

Summary Equations presented in this paper represent the steamdrive process assuming steam rises quickly to an impermeable overburden. Subsequent growth of the volume heated results from increase in both the area and thickness of the region heated to steam temperature. Equations derived predictthe rate of increase of steam zone thickness and areal extent,the volume of oil displaced from the steam zone and the heated reservoir beneath it,the reduced injection rate that will sustain steam after an area has been headted to steam temperature, andthe additional oil displaced after steam injection is stopped. Predictions are compared to observed behavior of Chevron's Kern River (CA) 10-pattern pilot steamdrive. Introduction Marx and Langenheim1 presented a valid relationship between the rate of injection of a hot fluid and the area heated by the fluid. The usefulness of the resulting predictions is limited by the assumption that hot fluids heat each areal increment of the reservoir uniformly from top to bottom. Typical steamdrive performance is different from that predicted by Marx and Langenheim because the difference between densities of steam and liquids causes steam to rise to the top of a permeable interval. Because of this gravity override, heat losses are not to the confining impermeable beds but partly to the permeable reservoir beneath the steam zone. Several models based on assumed or observed average steam zone thickness have been proposed. The model presented here enables prediction of steam zone thickness as well as the area covered by steam. Conditions Assumed Steam is assumed to rise to the top of a permeable reservoir in time negligible compared with that required to heat the total reservoir area. The horizontal pressure gradient in the steam zone is assumed to be much less than the vertical pressure gradient on liquids caused by the difference between densities of steam and liquids. Single values are assumed for both oil and water saturations throughout the steam zone. Fortunately, predictions depend only indirectly on water saturation, which seldom is known accurately. Because the horizontal pressure gradient in the heated reservoir immediately below the steam zone equals the small horizontal pressure gradient in the steam zone, horizontal flow immediately below the steam interface is neglected. Oil saturations in the heated reservoir beneath the steam are assumed to be the residual oil saturation to hot-waterflooding. The thickness of the interval flooded by hot water is represented by a single temperature that be estimated from laboratory data or field measurements. With this assumption, temperature-dependent relative permeabilities and flow properties are represented by a single temperature and a single oil saturation. The portion of injected heat that leads to steam zone growth is assumed equal to the portion in steam vapor. Injected hot water, like that from steam condensation, is assumed to fall quickly to the reservoir beneath the steam zone, where it contributes to oil displacement but does not influence steam zone growth. Predicted Steam Zone Growth and Oil Displacement Equations representing the concepts described as follows are presented in the Appendices. The rate of steam condensation resulting from heat diffusion into the overburden is a function of the time since the areal increment considered was first heated (Appendix A, Eq. A-4). The volume rate at which water flows downward through the steam zone equals the rate of steam condensation at the overburden/reservoir interface.The volume rate (per unit area) of water flowing downward beneath the steam interface minus that in the steam zone equals the volume rate of water displaced by movement of the steam interface (Eq. A-5).The enthalpy (per unit area) carried by the downward flow of oil and water beneath the steam interface minus that carried by water flowing downward through the steam zone equals the rate of interface motion times the difference in enthalpy above and below the interface (Eq. A-7).Vertical displacement of the steam interface does not require a change in the temperature gradient just below the interface.

The bitumen contained in Axe Lake Discovery in Saskatchewan has very high viscosity, well over 10 million centipoises under original reservoir condition. Since bitumen viscosity decreases with increasing temperature even at low temperatures, and since the Axe Lake reservoir is highly permeable and extremely coarse grained, preliminary reservoir simulations indicated that it is possible that effective, early time, communication between production wells at the bottom of the reservoir can be established by efficient low energy heating and intermittent water pressure pulsing. In order to evaluate the heating concept, we measured the response of the reservoir to heating with down-hole heaters at low temperatures (100 °C). A description of the field test is provided. Different analytical formulations and simulation models were used to analyze the results of the field test and the analyses were compared to prior published work. Field results were also used to determine effective thermal conductivities and effective mobilities at low temperatures for different grid sizes using Cartesian, radial and hybrid simulation grids. A comparison to analytical heat transfer formulations under different boundary conditions is discussed. Calibrated simulations show that not only can the heater be used to preheat the formation but also as an efficient bitumen viscosity reduction tool to increase bitumen mobility during the full bitumen production life of the reservoir. Moreover, high temperature heaters can be combined with cold water injection to generate steam in the reservoir directly. The paper concludes with a simple analysis which shows that such a combined technology can produce bitumen more efficiently and economically when compared to conventional SAGD. Introduction Axe Lake Discovery is located in northwestern Saskatchewan, about 90 km northeast of Fort McMurray. The bitumen reservoirs are buried about 200m below the surface. The bitumen is quite heavy and highly viscous, with a viscosity of over 10 million centipoises at original reservoir condition. Since bitumen viscosity decreases when the temperature rises, thermal technologies such as steam and hot water injection have become the preferred means for bitumen recovery. However, the bitumen reservoir at Axe Lake is relatively shallow and the efficient recovery of bitumen may be not possible using high pressure steam and hot water injection. Preliminary reservoir simulations indicated that it is possible to establish effective communication between producers at early time by efficient low energy heating and intermittent water pressure pulsing. Measuring the response of the reservoir to heating with downhole heaters at low temperature 100 °C, a two-stage test was planned and the second stage is currently being implemented. Different analytical formulations and simulation models were used to analyze the results of the measurement. The effective thermal conductivity and mobility of the bitumen at low temperature will be obtained from the results. Simulation studies also indicated that it was possible to obtain bitumen recovery effectively and economically using high temperature down-hole heater combined with cold water injection. This method takes advantage of simple and safe surface facilities and without losing heat outside the wells when compared to conventional steam injection from the surface.

Casing temperatures and wellbore heat losses are critical variables in steam and hot water injection wells. Several papers have been written presenting methods of estimating these parameters if the over-all heat transfer coefficient is known. The over-all heat transfer coefficient for a wellbore is developed from its component terms to promote a better understanding of the concept. Specific methods have been selected from the heat transfer literature for estimating the size of each heat transfer component. Simplified calculation procedures are suggested for determining the over-all heat transfer coefficient. Comparison of calculated and measured casing temperatures during steam injection confirms the basic formulation and applicability of the suggested procedure for engineering calculations. Introduction The design of steam and hot water injection projects requires estimation of casing temperatures and wellbore heat loses. Several authors have shown that wellbore heat losses and casing temperatures can be calculated if the over-all heat transfer coefficient is known. This article discusses methods of determining the over-all heat transfer coefficient from the process variables. Development The steady-state rate of heat flow through a wellbore Q Btu/hour is proportional to the temperature difference between the fluid and the formation, and the cross-sectional area perpendicular to the direction of heat flow. The proportionally factor, called the over-all heat transfer coefficient, represents the net resistance of the flowing fluid, tubing, casing annulus, casing wall and cement sheath to the flow of heat. Thus, we can write Q = Uj Aj Tj............................ (1) Eq. 1 defines Uj, the over-all heat transfer coefficient based on the characteristic area A, and a characteristic temperature difference T,. Subscript j in Eq. 1 identifies the surface upon which these quantities are based. In theory, any radial surface could be used to determine the characteristic area. Some choices are more convenient to work with than others. For example, if hot fluid is injected down tubing it is preferred to let A, be the outside surface area of an incremental length of injection tubing, 2 to L, and let T, be the difference between the temperature of the flowing fluid Tj and the temperature at the cement-formation interface (the drill hole) Th. Then Uj = Uo, referring to the outside tubing surface area, and Eq. 1 would be Q = 2 to Uto (Tj- Th) L................ (2) If the fluid is injected down the casing or casing annulus. the characteristic area could be the inside surface area of the casing, and Eq. 1 would be written as Q = 2 ci Uci (Tj - Th) L..................(3) Subscript ci refers to the inside casing surface. An expression for the over-all heat transfer coefficient for any well completion can be found by considering the heat transfer mechanisms between the flowing fluid and the cement-formation inter-face. A brief derivation of the over-all heat transfer coefficient is presented in the following paragraphs for the case of a hot fluid flowing through tubing insulated with a dry air annulus. Other cases can be derived easily once the basic concepts are understood. Fig. 1 shows the wellbore model which will be used to derive Uto. Heat Transfer Mechanisms The rate of heat transfer between the flowing fluid and inside tubing wall is given by Eq. 4. Q = 2 ti hj (Tj - Tti) L...................(4) hf, is defined by Eq. 4 and is the film coefficient for heat transfer based on the inside surface area of the tubing (subscript ti) and the temperature difference between the flowing fluid and the inside tubing wall Tj-Tti. Heat flow through the tubing wall, casing wall and the cement sheath occurs by conduction. Fourier discovered that the rate of heat flow through a body is directly proportional to the temperature gradient in the medium. JPT P. 607ˆ

A fast simple method is given for calculating heat losses to cap and base rock in thermal reservoir simulators. The method can handle backflow of heat and cyclic temperature variations. Test comparisons against analytic results are given. The method gives remarkably accurate results on the test problems, considering its simplicity. Introduction In thermal reservoir simulations, it is necessary to calculate the heat loss from the reservoir to the surrounding environment, usually to the cap and base rock. A finite-difference calculation can become expensive, especially on computer storage, as the temperature profile can extend large distances into the overburden and underburden, requiring large numbers of grid blocks for its description. However, high accuracy in the solution of the heat loss calculations is not usually required in reservoir simulators owing to the uncertainty in cap and base rock thermal parameters. This paper describes an efficient semi-analytical method for the case where the one-dimensional linear heat conduction equation is adequate for the heat loss calculations. Semi-analytical methods have been used previously (1,2,3), and a study of the results given in these papers leads us to the following tentative conclusions:Conduction within the cap rock rapidly wipes out any sharp differences in temperature. Thus, the temperature in the cap rock varies smoothly, even for rather erratic changes at the interface.Longitudinal heat conduction in the cap rock can usually be neglected, because the Peclet number is normally high for reservoir floods(2).The most important physical features of the heat loss process are good descriptions of the boundary conditions at the interface and conservation of cap rock energy. The tail of the temperature distribution contains little energy and is relatively unimportant. These points suggest that the temperature profile into the cap or base rock can be adequately approximated by any reasonably flexible function containing a few parameters. If such a function could be made to fit the thermal diffusivity equation with appropriate boundary conditions, and the calculation of the parameters could be done in a simple fast fashion, the method would become extremely attractive further justification for such an approach is that high accuracy is not required, as the coefficient of thermal conduction of the cap and base rock is rarely known precisely. The Method We choose as the fitting function for the temperature profile into the cap or base rock: (Equation Available In Full Paper) θ is the temperature at the interface between the reservoir and the cap or base rock (the zero level has been defined at the initial interface temperature); p and q are the fitting parameters yet to be determined; d may be interpreted as the diffusion length and should be of order √kt where t is the time measured from the instant at which the interface temperature first begins to change. The deepest penetrating term in (1) is z2e−z/d, which has its maximum at z = 2d. Hence we choose for the diffusion length:

Improving Steam-Assisted Gravity Drainage performance in oil sands with a top water zone using polymer injection and the fishbone well pattern

Thermal recovery processes such as cyclic steam stimulation and steam assisted gravity drainage involve large volume fluid and steam injection into oil sand reservoirs. Thermal expansion and dilatation of oil sand formation causes horizontal and vertical displacements in the overburden. These deformations may not exert detrimental effects on casings and facilities if the deformation properties of the overburden are homogeneous without drastic local variation along depth. However, if there exists some natural (clay seams) or induced weak layers (swollen and softened shale layer), shear slip could develop along these discontinuities, thereby impairing the nearby casings and facilities. This paper develops a semi-empirical analytical method for estimating the relative shear slip along a weak layer in shale due to fluid and steam injection. The magnitude of slip depends on the volume and distribution of fluid and steam injection, and the extent and location of the weak layer. Calculated results demonstrate that the shear slip along the weak layer could impose concerns regarding the steaming strategy and environmental protection.

Enhancing heavy oil mobility and drainage induced by reduction of viscosity and density due to temperature increase caused by steam injection is one of the most used methods by the industry. In spite of the vast investigation done in literature for the area of steam injection, the mechanism(s) of heat flow in reservoirs are not fully understood. Understanding heat flow mechanism(s) and temperature distribution, contribute to better optimization of the used steam. It is attempted in this paper to shed light on thermal recovery and temperature distribution with steam injection from experimental work. Basically, a model reservoir is constructed with heavy oil to be recovered by steam injection. Mapping of temperature propagation within the model reservoir clearly demonstrates fast steam/ temperature propagation to the overburden and then propagate horizontally at the top of reservoir. Hence, top of the reservoir becomes the second heating source of the underlying layers. However, the obtained temperature propagation profiles of simulated heavy oil thermal recovery by steam are different from the experimental data. Analysis of the temperature distribution shows that at the interface between the heated and cold region of the reservoir are initially contributed by convective heat flow (steam condensate and mobile heavy oil), then conductive heat transfer dominates. As the temperatures reach / close to steady state in the different heated areas, maximum drainage process takes place. Different simulation and experimental temperature distribution data are compared and showed better agreement between the estimated and experimental data as a function of time, indicating more conductive heat transfer mechanism with time. It is interesting to observe that in spite of the complexity of different processes taking place, consistent temperature gradients are established within the reservoir. This could lead to a better optimization of steam injection regime.

A steam injection test, carried out in the 1.5 m physical simulator developed at the Alberta Research Council for the evaluation of in situ recovery processes for oil sands is described. The experimental facilities and procedures, together with the test data obtained are described. Numerical modeling studies performed to simulate the test run using an implicit steam model are outlined. The capabilities of the numerical model, its mathematical formulation are described briefly without using equations. Relevant data for the simulation are given and the experimental results are compared with the model predictions. Introduction Physical and mathematical models are used to obtain 1n understanding of the recovery processes of bitumen from oil sands by steam injection. Both these approaches, in principle, can be used to extrapolate laboratory data to the field and predict the performance of a reservoir. Physical modeling, using scaling factors obtained from dimensional and inspectional analysis, even though desirable, is in fact difficult because, not all relevent mechanisms can be scaled. Mathematical models, on the other hand can be used effectively provided they have been validated by experiments. These experiments can be performed on reservoir elements or elemental models1. The materials and operating conditions in these elemental models are such that they represent some point in the actual reservoir and the recovery process involves the same mechanisms that operate in the reservoir. The physical shape and side of the elements is determined by the constraints of the test equipment. In the following sections, the experimental equipment at the Alberta Research Council, to studythe in situ recovery of bitumen from oil sands by steam injection, is briefly described. This equipment, generally known as the 1.5 m physical simulator facility, can be used for scaled models as well as elemental models. The numerical modelswhich have been developed at the Alberta Research Council to simulate steam floods in the physical simulator are introduced. A steam injection testcarried out in the 1,5 m physical simulator and its numerical simulation using an implicit steam model are discussed. EXPERIMENTAL EQUIPMENT A schematic diagram is shown in Figures 1 and 2. The equipment consists of a high pressure vessel, which contains the oil sand bed, three injection systems (steam, solvent and gas) and equipment to handle the production fluids. Process control and data collection is handled by a FOX 2/10, Foxboro minicomputer. The Physical Simulator The physical simulator is designed for a maximum operating pressure of 11.4 MPa. It has an internal diameter of 1.5 m with an overall height of 2.5 m. The vessel is separated into two parts by a flange. The lower part which contains the oil sand is 1.5 m high. The upper part is pressurized with nitrogen gas to simulate the overburden pressure. The nitrogen gas is separated from the oil sand bed by means of a deformable carbon steel plate of thickness L2 AWG, which transfers the overburden pressure to the oil sand bed.

EOR operations involving steam injection lead to heat being transferred to the reservoir, which expands, and the expansion is transferred to the overburden. This can induce a cap rock breach and steam/gas seep, leading to inefficient development and higher recovery costs. Thermal expansion could also re-activate faults, potentially affecting surface facilities. In a forward-modeling study of the cap rock integrity during steam injection in a giant fractured carbonate oil field in Oman the possible failure mechanisms of the cap rock were illustrated. Further, opportunities for cap rock and reservoir surveillance were identified. During a first phase steam development of the field a variety of well and reservoir surveillance data are being gathered, including (GPS and InSAR) surface deformation measurements, micro-seismic activity, and (distributed) temperature and saturation logs. Integrated usage of the surveillance data allows for calibration of and comparison between two geomechanical models. The geomechanical models are employed to monitor the cap rock integrity during the steam operation on a regular basis, and to derive the heat distribution (evolution) within the reservoir. To this end, both forward and inverse modeling workflows have been developed. The results of this study may trigger the development and evaluation of alternative steam injection schemes (well spacing, injection rates).

Summary Fireflooding of tar sand reservoirs following a preheating phase exploiting a reservoir heterogeneity was investigated in three dimensional (3D) physical models. The research strategy for applying the results and the experimental facility developed for this work are discussed along with the results from four tests. Three different types of reservoir heterogeneities were used: communicating and noncommunicating bottom-water zones and a thin, heated layer simulating a fracture. The preheating phase with the bottomwater zone was found to have a pronounced effect on fireflood performance. A short steam preheating phase did not provide sufficient oil saturation in the communicating bot-tomwater zone to ensure formation of a closed combustion front. A longer steamflood at a lower rate resulted in gravity override of the steam and the subsequent fireflood. A thin, electrically heated horizontal layer simulating a heated fracture yielded encouraging results with potential for further improvements. Introduction The great demand for crude oil and the decrease in conventional reserves have generated more interest in heavy oil and tar sands. Although an important fraction of the world crude supply comes from heavy oil, there is nocommercial in-situ production of the immobile oil con-tained in tar sands. However, tar sands have become an attractive alternative source for liquid hydrocarbons because of the size of the resource base. Canada and Venezuela have the largest tar sand/heavy oil deposits, each estimated at more than 1012 bbl (150×109 m3) of hydrocarbons in place.1x002D;3 Currently, most of this resource is too deep to be recovered economically by surface mining. Extensive work is being performed to develop in-situ methods for producing tar sands.2x002D;4 Although the sands are usually very permeable, the bitumen is immobile because of its high viscosity, often exceeding 106 cp (1 kPa·s). Most of the proposed recovery methods apply heat to the reservoir to increase the bitumen temperature, which drastically reduces its viscosity. Steam injection has been the most common method used for heating the reservoirs. However, fireflooding could be performed with more effective use of surface energy.5 In-situ combustion has been tested in more than 100 field applications, many of which have been reviewed and compared in the literature.6x002D;8 Screening guides have been developed for evaluating the suitability of reservoirs for forward combustion.7,9,10 An important criterion in each of these guides is that the reservoir must have sufficient fluid transmissibility. Since most tar sand reservoirs do not meet this criterion, they traditionally are not considered good candidates for fireflooding. Local heating of the reservoir near an injection well will not provide the communication between wellbores necessary for a displacement process. In-situ combustion in a tar sand requires some scheme to provide initial interweIl communication and mobility to a portion of the tar. A pretreatment might involveheating through a naturally occurring reservoir heterogeneity, such as an adjacent mobile water zone, or an induced fracture with steam injection or fireflooding. Combustion has been field tested in induced frac-tures11,12 and at oil/water interfaces.13 Alternatively, interwell communication might be achieved by reverse combustion14 or electrolinking. Preheating also could be performed by extensive steam stimulation or electrical techniques. The cost of the preheating phase might limit its duration to the minimum extent necessary to begin a displacement process.