Abstract Relative permeabilities are complex and important rockfluid properties of reservoirs in which multi-phase flow conditions prevail. Measuring relative permeabilities in the laboratory using cores obtained from a reservoir is a complicated, time demanding and labor-intensive task. There has been limited success in mathematical modeling of relative permeabilities based on rocks and fluid properties due to our inability to simulate the non-linear controlling mechanisms in place. Artificial neural networks (ANNs) promise a potential avenue for implicitly incorporating the controlling mechanisms and parameters into a model that can be utilized as an effective tool for relative permeability predictions. The methodology described in this paper exploits the unique topology of ANNs for determining the two-phase (oilwater) relative permeabilities. The ANN is a universal approximator that performs non-linear, multi-dimensional interpolations. in the development stage of the ANN model, a large number of oil-water relative permeability data sets were collected from the literature. These data sets were used to train the model. In composing the architecture of the ANN, only the readily available rock and fluid properties (endpoint saturations, porosity, permeability, viscosity, and interfacial tension) have been explicitly incorporated. The predictive ability of the model was tested using experimental data sets that were not used during the training stage. The results are in good agreement with the experimentally reported data. The proposed model exhibits sensitivity to several reservoir properties. The proposed ANN model has a dynamic training base that can be expanded as new data become available. Introduction Relative permeabilities quantify multi-phase flow through porous media. Generating an accurate relative permeability versus saturation relationship for each phase is essential for evaluating the performance of a reservoir during primary, secondary, and tertiary production periods. Relative permeability-saturation relationships vary between reservoirs and within a given reservoir. They are non-linear functions of reservoir rock and fluid properties such as phase saturations, formation types, depositional environment, shale content, heterogeneity, porosity, permeability, interconnectivity of pores, pore geometry, interfacial tensions between flowing phases, phase viscosities, phase densities, and rock wettability. Experimental and modeling methods are used for assigning relative permeabilities to a reservoir. Although laboratory measurements of relative permeabilities are difficult, they are still the preferred method. Laboratory measurements are technically difficult and require skillful personnel, expensive equipment, and are lengthy to perform. Therefore, estimation of relative permeabilities with mathematical models has always been an attractive goal. The accuracy of relative permeability values must be preserved. Relative permeability models are commonly used only as estimation tools. Mathematical models for relative permeability can be classified under four main categories: capillary, statistical, empirical, and network models. These models require some rock and fluid properties such as endpoint saturation values, porosity, absolute permeability, interfacial tensions, and viscosity of phases, and incorporate important assumptions. The models are restricted by their assumptions, are not universally applicable, and may be difficult to update for different systems. Empirical models and pore-network models are frequently used and are the most successful in estimating relative permeabilities. Predicting the relative permeability values using mathematical models is

Abstract The effects of permeability contrast in heterogeneous porous media and the role played by pore level physics on the production characteristics of oil and water in the gravity assisted inert gas injection (GAIGl) process were studied experimentally using I) long columns of sand packs. 2) columns of glass beads, and 3) consolidated models of glass beads between glass plates for flow visualizations. The percent oil recovery efficiency in stratified(layered permeability type) porous media under water wetting and positive oil spreading coefficient conditions was found to be about 80 % of the waterflood residual oil saturation or better. No significant amount of oil is held at the boundaries of the low to high permeability zones. The main reason for this is the presence of a continuous oil film over water in the edges of pores invaded by the gas throughout the medium which permits the oil to continuously drain towards the production end at the bottom of the system. A mathematical analysis of the GAIGI process has been developed that can be used to predict the production history and oil saturation profiles during gravity drainage for both, homogeneous and layered permeability type of porous media. Excellent agreement was found between the experimentally measured oil production characteristics and those predicted theoretically. Introduction The gravity assisted inert gas injection process (GAIGI) or tertiary immiscible gas flooding has been the subject of three PhD studies in our laboratory (Kantzas(l988), Catalan(l992), Ayatollahi(I994?) and related publications, as well as, of others who were influenced by our pioneering work on this very promising tertiary oil recovery process(e.g., Oren et. al. (1992), Pavone(1989). Vizika(1990?). Very high oil recovery of residual oil is attainable if gravity drainage (the self propulsion of oil downwards in a reservoir) is the dominant production mechanism. A significant contribution of recent laboratory studies on tertiary immiscible gas injection schemes has been in the development of a better understanding of the pore level physics of immiscible displacements involving three immiscible fluids, namely, oil, water, and gas (Chatzis et. al(1988), Kantzas et. al(l988)). The role played by pore structure(Catalan et. aL(l994), Kantzas et aL(l993), Oren and Pinczewski(1991)), the wettability(Chatzis et. al.1988), Catalan el. aL(1994)) and 'the effects of oil spreading coefficient on oil recovery, respectively, have also been investigated for the GAIGI process(Chatzis et, al(l988), Catalan et al(1994), Vizika(1993)). The underlying principle of the tertiary immiscible gas flooding process is the formation of a gravity stabilized oil bank from the residual oil in the course of gas invasion into pores containing residual oil and water. The most common conditions that have been investigated in depth are those corresponding to water-wet media and positive oil spreading coefficients. Under these conditions, when pores filled with waterflood residual oil are contacted by a gas, after dented by the gas, the oil in the blobs spreads spontaneously over the water and forms bulk films of oil between the connate water and the displacing gas which is at displacement pressure.

Abstract The objective of this paper is to evaluate experimentally, the effect of temperature on the three-phase brine-oil-gas relative permeabilities of an unconsolidated sand system. Two and three-phase relative permeabilities have been measured at 75 °C and 125 °C. The experiments were performed under steady-slate conditions. These measurements were conducted in Ottawa sand using a clean mineral oil, 1% brine and nitrogen gas. The measured relative permeabilities showed no significant temperature effect. The three-phase brine relative permeability was found to be a function of brine saturation alone. The three-phase gas relative permeability was always lower than the two-phase values. This is consistent with some of the previously published results. The oil relative permeability was found to vary with the saturations of the other fluids. Oil isoperms were concave towards the oil apex. Filially, oil relative permeabilities were compared with predictions obtained using modifications of Stone's method I and II. The prediction of Stone I was good for both temperatures. But Stone II predictions were consistently lower than the experimental values. Introduction Three-phase relative permeability plays an important role in the numerical simulation of oil recovery processes. Thermal methods for heavy oil recovery, such as steam drive and in-situ combustion involve the simultaneous flow of three immiscible fluids through the porous rock formation at elevated temperatures. Over the past 25 years, a number of studies have reported temperature effects on two-phase relative permeabilities in porous media. Several of these 1–6 report a significant effect of temperature on oil/water relative permeability. Although it is probable that three-phase relative permeabilities would also be temperature sensitive, no experimental study of the effect of temperature on three phase relative permeability has been reported. Whether or not three-phase relative permeabilities are temperature sensitive for a given system can only be determined by experimental evaluations. The objective of this work was to evaluate experimentally whether the three-phase relative permeabilities are temperature sensitive for a given system or not. The experiments were performed under steady-state conditions, thus minimizing instability phenomena. and allowing for control of saturation history. The relative permeability for each fluid was calculated by applying the following form of Darcy's law to each fluid: Equation (1) (Available in full paper) All the variables in Equation 1 (other than q, and ΔP 1 can be measured separately. Because the usual objective of such measurements is to determine the relationship between relative permeability and saturation, it was also necessary to measure the saturation of each fluid under steady-state conditions for each set of flow rates. Oil saturation was measured by the material balance method. Brine saturation was measured by the electrical resistivity method. A calibration curve, relating the electrical resistivity index of the core to its brine saturation, was established by carrying out two phase oil/brine and gas/brine tests. The resistivity index was measured directly by measuring the ratio of the electrical resistance of the core and the resistance of the fully brine saturated secondary cell.

Abstract In situ recovery of heavy oil and bitumen from oil sands often involves simultaneous flow of oil, water and gas in the oil sands. Such multi-phase flow in porous media is a complex process. Mathematical simulation of such recovery processes requires a knowledge of the threephase relative permeability characteristics of the system. Experimentally measured three-phase relative permeability data for heavy oil reservoirs are not available ill the literature. This lack of data is primarily due to the tedious nature of such measurements and shows a need for developing less time consuming methods. In this study, an unsteady-state technique similar to the JBN method for two-phase flow was developed and used to estimate three-phase relative permeability data. The method was validated by comparing the oil isoperms with those obtained using the steady-state method. A reasonably good agreement was obtained which suggests that the proposed unsteady-state method could be employed as a faster alternative to obtain three-phase relative permeability data. Introduction Petroleum production often involves simultaneous flow of three immiscible fluids through subterranean porous rock formations. Examples of recovery processes involving such three-phase flow include; alternating gas and water injection, steam drive, in-situ combustion, and immiscible gas injection in presence of mobile water. Thorough analysis of these processes is impossible without reliable three-phase relative permeability data. Reservoir simulation programs incorporate one or several techniques for estimating three-phase relative permeabilities, which permit three-phase simulation studies to be carried out with input of only two-phase relative permeabilities. However such simulation studies may be no more reliable than the techniques for estimating three-phase relative permeabilities employed in them. Unfortunately, the reliability of these estimation techniques can not be tested without availability of extensive experimental data on three-phase relative permeabilities Compared to the wealth of information available in the literature on two-phase relative permeabilities, there is a dearth of experimental studies on three-phase flow through porous media. Virtually no information is available on three-phase relative permeabilities in intennediately-wet or oil-wet systems. Effects of variables such as: contact angles, viscosity ratios. interfacial tension, flood velocity and temperature on three-phase relative permeability have not been investigated. There is an obvious need for more experimental investigations of threephase relative permeabilities to resolve some of the uncertainties concerning the effects of variables mentioned above. The reason for this scarcity of experimental data on three-phase relative permeabilities appears to be the effort and time required for such measurements. The only well accepted technique for such measurements is the steady-state method. Between 50 to 100 steady-stale measurements are required to fully define the relative permeability characteristics of a given system for one direction of saturation change. It often takes 24 hours or more to reach steady state conditions for each such measurements. Combined with the time involved in core preparation and periodic extractions to overcome the problem of error accumulation the total effort involved is indeed formidable. If the hysteresis effects corresponding to various possible directions of saturation change are also to be evaluated, the task becomes even more onerous.

Abstract In situ combustion, and its modified forms, continue to be important oil recovery processes. The applicability of in situ combustion to a wide range of reservoirs, and the recent advances, show that this process has unique advantages (and disadvantages). Laboratory studies of in situ combustion are hampered by the lack of suitable scaling criteria for the design of experiments. Apart from that, the experiments are usually labour-intensive and costly This paper provides useful guidelines for the design of such experiments, and also discusses the laboratory methods currently used. The general problem of in situ combustion in a three-dimensional system is formulated, with attention to reaction kinetics and non-equilibriwn phenomena. Appropriate boundary conditions, both for field and laboratory, are stated. Next, the partial differential equations are used to derive a set of scaling criteria. A number of alternative schemes are considered in this regard. Suggestions are offered concerning the use of the various options available for experimental design. The scaling criteria derived are. used to examine the validity of results derived from "fire tube tests, which are widely used to obtain basic information on the combustion bevaviour of crude oils. It is shown that among the parameters measured in such experiments, only a few, such as fuel content, have validity. Results of laboratory fire tube tests of wet combustion can be misleading. Introduction The in situ combustion process has been a source of interest, effort, and, at times, frustration for at least 30 years. The concept of burning heavy fractions of the reservoir oil in order to generate heat in situ is attractive and consequently has led to over 100 field tests of such processes [40 listed in Chu and Crawford 1 ]. However, due to very unusual complications only a few of these have reached commercial status. Even though operational problems have been pointed out as the major cause of such economic failures, a more subtle reason is the selection of reservoirs for in situ combustion applications. Selection criteria used in the past appear to be inadequate in the face of growing recognition of the complexity of the mechanisms involved. Although there has been extensive laboratory study of most aspects of in situ combustion, there remain several areas which are poorly understood. The objective of this paper is to elucidate the mechanisms involved in in situ combustion processes and to present various scaling options in order to model such a process. Mechanisms of In Situ Combustion Recently Islam, Chakma and Farouq Ali 2 revealed the mechanisms involved in in situ combustion and proposed a new mathematical model in order to model some extraordinary features which have been largely ignored by previous researchers. In situ combustion is based on burning the coke deposit laid down by the crude oil. The mechanism of in situ combustion varies depending on whether forward or reverse combustion is taking place. In all cases, however, oxygen containing gas (most commonly air) is injected into the reservoir to achieve combustion.

Abstract When making a waterflood prediction, one of the more important properties needed is the water-oil relative permeability characteristic of the reservoir rock. Often the relative permeability correlation for a particular rock-fluid system is established on the basis of laboratory displacement studies. Moreover, the external-drive technique are truly representative of the actual flow phenomena taking place within a core. The purpose of this study was to further validate the external-drive method. This was accomplished by measuring the saturation and pressure gradient at the outlet end of a core and comparing these values with those estimated using external-drive theory. Good agreement was found between the values determined directly and those predicted using the external-drive theory. Moreover, good agreement was found between the dynamic experiment and the equilibrium capillary pressure curve determined using the same sand-fluid system. Finally, the measured saturation profiles were in good agreement with those predicted using Buckley-Leverett theory, provided th displacement was stable and stabilized. Introduction The water-oil relative permeability characteristics of a reservoir rock is one of the more important properties needed to undertake a waterflood prediction. Either steady-state of external-drive techniques may be used to obtain such information. Because of the long time needed to obtain stabilization when using steady-state methods, the external-drive(unsteady-state) methods are usually preferred. Moreover, the unsteady-state methods are applicable to displacements where saturation gradients exist, while the steady-state methods are not. The unsteady-state methods are based on Huckley-Leverett theory 1, . As a consequence, displacements used to measure relative permeability must be conducted in systems which are linear and homogeneous. Moreover, it must be permissible to neglect the capillary pressure gradient. In addition, the displacement must be stabilized and one dimensional (pressure and saturation uniform in any cross-section 2 ). Also, a Lagrangian formulation of the fluid displacement problem must be permissible. That is, it is important that the flow is neither unstabilized nor unstable, and that the saturation profiles are monotonic, when using external-drive methods. 3 In the past, it has been usual to validate the external-drive techniques by comparing the relative permeability curves obtained using these techniques with those obtained using steady-state methods. 4,5 . Good agreement has been found between the curves derived using these two approaches, provided the assumptions underlying the external-drive technique are well met. However, if one of the underlying assumptions is seriously violated, problems can arise. For example, if the sample used for the displacement tests is heterogeneous, the relative permeability curves obtained using the external-drive techniques can differ significantly from those obtained using other techniques 4,6 . Recently, a new approach to measuring relative permeability has been under development at the University of Alberta 7,8 . As a consequence, it has become possible to verify more directly the assumptions and theory underlying the external-drive method. For example, because it is possible to measure directly the saturation at the outlet end of the system, it is practicable to compare the measured value with that estimated using the equation developed by Welge 9

Abstract A water alternating gas project was started in late 1985, south of Lloydminster in section 06-049-27 W3M. Waterflooding, as well as water alternating gas injection, have been evaluated in the laboratory. These tests demonstrated that a methane WAG process would be effective in increasing the oil recovery in section 06-049-27 W3M. Recovery over primary is predicted to be 5.5%. Considering the thin net pays of the Sparky "B" and lie" zones (approximately 3 meters each), the underlying water in Sparky "C" zone and the existence of a shale barrier, thermal recovery techniques were not feasible in this location. Introduction The Silverdale Sparky Pool is located in the Lloydminster area of Western Saskatchewan and Was discovered in 1945 (see Figure 1). Development drilling of section 06-049-27 W3M began in mid-1977. Section 06-049-27 W3M is 100% owned and operated by Home Oil Company Limited. The formation is a typical Sparky with good porosity (29%) and high oil saturation (82%). Before the implementation of the water alternating gas (WAG) scheme, primary production from section 06 was completely depleted. A WAG project has started in late 1985 and is composed of 3 patterns; two inverted 6-spots and one inverted 5-spot joined in the north-south direction. Low oil prices and a shortage of injection water have plagued the project over its first year of operation. It's too early to determine the effectiveness of WAG, but production is encouraging. GEOLOGICAL DESCRIPTION OF SECTION 06-049-27 W3M In section 06-049-27 W3M ~he producing horizon is the Sparky formation of Lower Cretaceous age. The formation is divided into two sand zones, Sparky Band C, separated by two thin shale beds. The average net pay of the Sparky Band C is 2.6m and 3.3 meters respectively. Figures 2 and 3 are structural contour maps of the Sparky Band C sand respectively. The structure rises slowly towards the east for approximately 2-1/2 LSD's, then is abruptly cut off due to a shale barrier runing north/south. The Sparky B sand has an additional geological feature, a shale channel runs diagonally NW/SE. Consequently the B zone is shaled out in a number of wells, notably 2B-6, 2C-6, 6A-6 and 6C-6. Two wells have been cored through the section (2C-6 and IOC-6). Above the B sand is a coal bed of approximately 1.5m in thickness. The coal discovered was black, brittle and of bituminous nature. The pay zones generally consisted of dark brown, poorly consolidated, fine to medium grain, round and well sorted sand grains with excellent porosity and permeability. Average water saturations for the Band C zones were 20% and 17.5% respectively. The shale interbeds encountered were generally medium gray to black, soft and non-calcareous, often containing minor sandstone lamina. A sieve analysis and x-ray diffraction analysis was carried out on a sample from 2C-6. The sieve analysis revealed that anywhere from 1%-2.4% of the sample was clay (less than 5 microns) material. The x-ray diffraction analysis showed that kaolinite, illite, chlorite and smectite were the only clays present.

Abstract Husky has extensive land holdings in the Lloydminster area, and owns additional properties in the Cold lake area. These lands contain oil fields which are shallow and relatively thin, but have good porosities. ail saturations, and permeabilities. Most of the reservoirs contain heavy oils, produce no more than 10% of the OOIP on primary production, and are good candidates for thermal EOR processes. Pilot locations were selected using screening criteria based initially on literature guidelines, but subsequently modified to reflect field results. Since 1965 Husky has conducted a wide variety of field tests using the cyclic steam, steamflood, and in-situ combustion processes. Both successes and failures have been encountered with each process, but the operating data and experience accumulated from earlier pilots have significantly improved the performance of the later pilots. There are current successful pilots using each thermal EOR method. A reduced effort has been made to study non-thermal EOR methods due to their limited benefit in heavy oil reservoirs. Immiscible CO, stimulation and electric stimulation have also been field tested on a limited basis. Introduction This paper was written to give the reader an overview of Husky's recent EOR pilot experiences in the Lloydminster and [old Lake Areas. Due to the number and scope of the projects discussed, the depth of technical detail has been purposely limited. If further information is desired, the reader is referred to the REFERENCES section. Reservoir geology has a strong influence on the app1icabi1ity of any EOR process. However, the subject is both beyond the depth intended for this paper, and the skill of the author. Its mention is only meant to remind the reader of another aspect of EOR technology development. The EOR projects discussed have been divided into two major types (Therma1 and Non-Therma1); and the Thermal EOR pilots have been further subdivided into cyclic steam, steam drive, and in-situ combustion. A Chronological sequence is followed within each division. Thermal processes have been emphasized by Husky due to their applicability to the Lloydminster and Cold Lake Reservoirs. General thermal EOR screening criteria used to match the appropriate process to each candidate reservoir are listed, but it should be noted that such criteria are revised whenever new technical or economic data become available. The material in this paper documents the evolution to larger, more sophisticated, and generally more successful thermal EOR pilots which has occurred at Husky (and throughout the industry). Unfortunately, less success has been experienced with thermal processes applied to thin (3 - 5 m) sands. Husky's efforts to develop non-thermal EOR methods are also discussed, although they are both fewer and smaller due to their apparent inapplicability to Lloydminster heavy oils reservoirs. GEOLOGY ANO RESERVOIR DESCRIPTIONS The Lloydminster heavy oil deposits farm the southern end of the discontinuous trend of Lower Cretaceous bitumen and heavy ail deposits which are also present in the Athabasca and Cold Lake areas. The Lloydminster deposits are unique in their ability to produce up to 10% of the OOIP under primary conditions.