Abstract The steady flow and viscoelastic properties of kerosene-inwater and mineral oil-in-water emulsions were investigated. using a controlled-stress rheometer. The rheological parameters (viscosity, storage modulus and loss modulus) show a substantial increase upon reduction of the droplet size of the emulsions. When fine and coarse emulsions having the same volume fraction of droplets (and the same oil) are mixed together in varying proportions, the resulting mixed emulsion exhibits a minimum in the viscosity, storage and loss moduli plots at a certain proportion of the fine droplets. However, the minimum in viscosity occurs only at low shear stresses. At high stresses. the viscosity of the mixed emulsion increases linearly with the increase in the fine emulsion content on a semi-log scale. Introduction An "emulsion" is a two-phase oil/water system where one of the phases is dispersed as droplets in the other. Emulsions also contain a third component. called the emulsifying agent or emulsifier. which has two principal functions: to decrease the interfacial tension between oil and water thereby enabling easier formation of the emulsion; to stabilize the droplets against coalescence once they are formed. Emulsions are encountered at nearly every step of the petroleum production and recovery operations viz. within the underground porous media (oil reservoir) at wellheads in phase separators in flotation units in crude oil transport facilities and at various stages of the refining process. Recently, considerable interest has been shown in the use of emulsions for pipeline transportation of highly viscous crude oils l . The transportation of extremely viscous crude oils. such as bitumen and heavy oils. by pipelines is difficult, especially during cold weather. To facilitate the flow of very viscous crude in pipelines. it is necessary to either heat the oil to reduce its viscosity or dilute the oil with a low viscosity hydrocarbon diluent. In the former case. heating cost and pipeline insulation costs are generally very high. In the latter case. the cost of diluent and its availability pose difficult problems. However. these problems can be avoided if the crude oil is transported in the form of an oil-in-water (O/W) emulsion. For emulsion pipelining to be economical. it is important to keep the oil concentration of the emulsion as high as possible. and at the same time. maintain the, emulsion viscosity at a reasonable level. It is believed thii(the key factors that determine the emulsion viscosity at high oil concentrations are the droplet size and the droplet size distribution. However. little research work has been carried out on the influence of droplet size and droplet size distribution upon emulsion rheology. In the past. numerous theoretical and numerical studies have been conducted on the motion/deformation of emulsion droplets and emulsion rheolog 2–41 . In the low-Reynolds number limit. emulsion rheology depends upon the following factors: volume fraction of the dispersed phase (Ø). ratio of the dispersed-to-continuous-phase viscosities (K) and the capillary number (N C1 ) The deformation of the droplets in shear flow causes normal stresses and shear-thinning behaviour in emulsions.

Abstract For over three years, a multi-disciplinary project team has been evaluating thebusiness opportunity of injecting carbon dioxide into a portion of the Weyburn Unit to improve recovery. As with any project of this magnitude, having fullcycle expenditures in excess of 2 billion dollars, a comprehensive analysis ofthe opportunities and risks in the project had to be clearly defined andevaluated. A risk assessment using a Monte Carlo simulation approach was undertaken tocombine the technical with the non-technical issues associated with the projectin order to define the full cycle opportunities and risks. The risk assessmentprocess was used to optimize the project configuration, to focus the team onkey issues, and to find ways to mitigate the risks inherent in theproject. Uncertainties impacting the range in project returns appeared in the areas ofcommodity prices, reservoir performance, costs and fiscal terms. The riskassessment process, as it was applied to the Weyburn CO 2 projectevaluation, is the subject of this paper. Introduction The Weyburn Unit, covering an area approximately 180 km 2 , isapproximately 130 km southeast of Regina, Saskatchewan (Figure 1). The Unit isoperated by PanCanadian Petroleum Limited on behalf of 40 working interestowners. The field was discovered in 1955 and produced under primary depletionuntil an inverted nine spot waterflood was implemented in 1964. Currentproduction averages 3150 m 3 /d of medium-sour crude from 650 oilproducers, of which 90 are horizontal wells. After waterflood, more than 116E 6 M 3 of oil will remain in the reservoir, or 65 % of theoriginal 178 E 6 M 3 in place, representing a very attractivetarget for an enhanced recovery process. The Midale beds of the Weyburn field were deposited on a shallow carbonateshelf in the Williston basin. The reservoir is unifonnly divided into the upper Marly, a chalky inter tidal dolostone with limestone interbeds, and a lowerVuggy zone, a heterogeneous and highly fractured subtidal limestone. A morecomplete reservoir characterization was discussed by Elsayed etal 1 . An analogy to the Weyburn field can be drawn from the Midale Unit, is situatedjust east of the Weybum field. Shell Canada Ltd., the operator of the Unit, implemented a CO 2 injection pilot in 1984 which was completed in1988. Due to encouraging results from the pilot 2 , Shell implementeda "CO 2 demonstration project", an eight pattern flood in 1991 whichwas designed to improve oil recovery from the area and provide technical andeconomic data to further assess full scale development potential. This flood isstill in progress. Input data forming the basis of this assessment incorporated some of the Midaleinformation as well as opinions from internal experts, industry consultants andother analogous CO 2 floods in the Permian Basin 3 Traditional evaluation process Traditional engineering, design, and estimating methodology sets out a logicalstep-by-step process which leads through the project life cycle.

Abstract The flow of surfactant-stabilized oil-in-water emulsions through wedge meters and segmental orifice meters was investigated experimentally. A flow loop consisting of wedge meters and segmental orifice meters of different shapes and sizes was designed and developed. The discharge coefficients were determined for various differently concentrated stable oil-in-water emulsions. The dispersed phase (oil) concentration of emulsions was varied from 0 to 76.14% by volume. At low to moderate values of the dispersed phase concentration, the emulsions were Newtonian in nature. At high oil concentrations of 65.16 and 76.14% by volume, the emulsions were non-Newtonian pseudoplastic in nature. The single phase calibration curves of discharge coefficient versus Reynolds number are found to be applicable to surfactant stabilized oil-ill-water emulsions. both Newtonian and non-Newtonian. However, in the latter case, one needs to use the generalized Reynolds number instead of the convential one. Based on the experimental data, empirical expressions for the wedge and orifice discharge coefficients are given. Introduction An emulsion is two-phase oil/water system where one of the phases is dispersed as droplets in the other. The phase which is present in the form of droplets is named "dispersed phase", and the phase which form the matrix in which theses droplets are suspended is referred to as the "continuous phase". Emulsions of oil and water are widely encountered in industrial applications. For example, a significant portion of the world's crude oil is produced in the form of emulsions. It is believed that the emulsification of oil and water occurs in the formation near the well bore where the velocity gradients are very high. The high shear in the production facilities such as pumps, valves, and turns also favors the formation of the emulsions [1]. The application of emulsions in industries other than petroleum are numerous; many products of commercial importance are sold in emulsion form. The non petroleum industries where emulsions are of considerable importance include food, medical and pharmaceutical, cosmetics, agriculture, explosives, polishes, leather, textile, bitumen, paints, lubricants, and polymer [1]. The measurement of flow rate of emulsions is required in many of the industrial processes where large volumes of emulsions are bandied. However, despite the industrial importance, little published work is available on metering of emulsions. It is due to this lack of information that the present work was undertaken. The major objective of the present work was to investigate the applicability of wedge meters and segmental orifice meters to emulsions metering. The wedge flow meters consists of a wedge-shaped constriction placed in the path of fluid flow to create a differential pressure as shown in Figure la. The restriction is characterized by the Z/D ratio and the wedge angle, where Z is the height of the restriction, and D is the pipe inside diameter. The segmental orifice meter (see Figure 1b) is a special case of a wedge meter having a wedge angle of zero. Wedge meters and segmental orifice meters have many applications in the Petrochemical industry and have many advantages as follows: (a) they can be used to measure the rate of flow in either direction,

Abstract Ten percent of Alberta's gas reserve is associated with low quality reservoirs known as the Milk River, Medicine Hat, and Second White Specks formations. Although the quality of the reservoirs is low the quantity of production information has has its own quality which provides confidence in historical type curves. These curves have been interpreted to describe the various productive characteristics of remaining reserve and available producability. Using the most recent information on productive rate, the cumulative production and the time oil production, two methods of type curve analysis will be compared. The problem of scaling the type curves to field measurements boils down to two equations in three variables. Our faith in our ability to plot a representative type curve provides the compatibility necessary to accomplish a meaningful scaling exercise. The first method will be dealt with briefly and outlines a graphical procedure which interpolates the relationship between the current rate and the initial tested rate of the well to determine the reserve associated for the well and field. The second and preferred method is more rigorously numerical. It involves averaging rate and cumulative derived indexes which provide a direct scaling factor for initial rare Will ultimate reserve. Direct comparison of the the two methods is provided at the hypothetical and practical levels using actual field data. The various characteristics for each formation are tabled and Gil example of a computer spreadsheet is incorporated. Introduction The quality of a shallow gas well is typically described in terms of its current rate, the volume produced and how old the well is. If this were the only reliable information available, determining the effectiveness of field improvements from sweet spots in the field would be a difficult and time consuming chore. Precise determination of shallow gas reserve has only limited practical application, since individual circumstances raise questions of whether a wells performance is related lo completion, well specific circumstance, or field management. This empirically consistent method will improve the grading characteristics for performance and rate projections. The following procedure has merit regarding the repeat able prediction of reserves and consequently the well's short term and ultimate producability. Comparisons of varying field management standards can then be achieved from the start point of reasonable ultimate reserves expectations. This procedure has broad use for the practical scaling of reserves on a well by well and field basis. It can also assist with consistency of production forecasting. The procedure is well suited to implementation through a simple spreadsheet which allows the summation of multiple wells and well groupings towards facility oplimiz3tion. The traditional method of type Curve analysis incorporates some type or initial rate testing and then relates the initial rate lo the Warren curves. (The curves contained in the paper titled Southeastern Alberta Shallow Gas Pools Reserves and Performance Data by Andy Warren, Energy Resources Conservation Board). When inconsistencies are experienced, questions regarding the quality of the initial test procedures and other data are raised. Since the initial testing is characteristically reflective or the operator in terms of consistency, the reserve estimates arc deemed to be approximate.

Abstract The type of R&D that is best suited to collaborative development has been a hot topic for several years. This paper develops a method to characterize R&D and collaborative groups that results in a simple analysis of collaboration decisions. The key descriptive parameter, for a particular company, for R&D is the ratio of internal to external potential benefits from the R&D if it is successful. This ratio is a function of the type of industry, the company's willingness and ability to achieve internal and external benefits and the nature of the R&D itself. The key parameters describing the collaborative group are the size of the group, the relative ability of an individual member to achieve external benefits in competition with his collaborators and the effectiveness of the group in conducting a R&D program. Conditions for collaboration are developed based on decision-tree analysis. The key requirement for effective R&D collaboration is that a company is capable of internalizing a Significant portion of the benefits. This determines good collaborators and good industries for collaboration. Collaboration benefits can also be improved by R&D exchanges in collaborative associations such as CONRAD. However, simple decision-tree analysis understates the advantages of collaboration. Portfolio risk reduction and inter-project synergies are significant additional advantages. R&D collaboration is the preferred route for development of a broad base of petroleum-related technologies. Introduction I have often heard that collaboration to develop new technology is only worthwhile for companies if the R&D is ‘pre-competitive’, with no end-use device or process in sight. Not only is the definition of ‘pre-competitive’ R&D vague and subjective, but the concept of limiting collaboration to this arena may be too restrictive. In fact, the incentive to collaborate is for much broader-based R&D, as will be described in this paper. In a general sense, there are a series of four logical, connected questions that are important for decisions oncorporate MD investments: Does the R&D prefect fit within the constraints of corporate strategy﹖ Do we want to invest in this area﹖ Is the risk low enough and the potential benefit great enough to economically justify an investment in an R&D prefect﹖ In other words, does the risk-weighted benefit/cost ratio pass the corporate investment hurdle test﹖ In the portfolio of potential investments, is the benefit/cost high enough so that budget funds are available? Does our budget prevent us from doing this prefect﹖ Given that the investment is justified, funds are available, and the prefect is of strategic interest, is it better to do the prefect oneself (go-it-alone) or to collaborate with peers, competitors and others to jointly develop the technology﹖ The purpose of this paper is to characterize R&D in a useful system for corporations so that the last question can be analyzed and the pros/cons of R&D collaboration can be discussed. Throughout the text it is assumed that: The R&D investment is attractive, funds are available, and it is of strategic interest Collaboration is not impeded by any legal or regulatory constraints.

Abstract In response to the drop In oil prices a decade ago. Husky Oil redesigned its EOR research management processes with the goal of increasing investment efficiency and reducing process development time. This paper describes that system. The sudden change in the crude price outlook dictated three key considerations for the new management process. The price change meant that recovery methods under development in the late seventies and early eighties were no longer applicable. A complete swift in target process threshold prices was required. Research had to be driven by the needs and constrains of the business. The need is to match the recovery processes to the reservoirs. Industry multi-company research programs were focused on processes that were not viable in the new price environment Also target reservoirs were not similar to the vast majority of the fields found in the Lloydminster area. The new R&D management system that resulted from Husky's redesign has the following key features Business questions are identified first. Then the technical questions that need to be answered to meet the business need are derived. Research is focused on answering these questions. Teams consisting of geologists, operations, reservoir and research engineers are used t0 screen processes and match them to reservoirs. EOR process ideas are generated from a clean slate with the business constraints of the day in mind. Introduction Husky Oil has a great deal of interest In the Lloydminster area. The company holds leases on much of the land and has spent a large amount of operating and capital money In the drilling and completion of wells. Also an extensive pipeline system, has been installed Lo carry the oil and water to various locations for upgrading and disposal respectively. There are approximately 12,000 wells In the Lloydminster area penetrating over 1,000 separate pools. Husky Oil owns approximately 4,000 wells. The majority of the wells penetrate into the over 45 regional or channel sands that have been recognized and designated by the Saskatchewan Energy and Mines (SEM). Total Husky controlled original oil-in-place is estimated at 0.7 × 10 9 m 3 . Over 40% of the oil-in-place in L1oydminster is contained in reservoirs less than seven meters thick (Miller, 1987) These reservoirs have been described as fine to very fine grained quartzose sand zones with porosities about 30% and permeabilites generally in the 0.5 to 10 Darcy range. Oils are to 10 ° API to 16 ° API with viscosity between 500 mPa.s and 50.000 mPa.s. A small fraction but a significant volume of the oil in place (about 50x10 6 m 3 ) is contained in channel sand deposits with thicknesses in the 10 m to 30 m range. The sand is typically located at a depth of about 400 m to 500 m at original pressures of about 3,000 kPa to 4,000 kPa and temperatures ranging from 20 °C to 30 ° C. The connate water saturation on the average is about 25%.

Abstract Effective resource management is required to assure safe And environmentally responsible disposal of fluid waste by injection wells and the development of oil and gas fields. Hydraulic fractures which break out through the upper confining stratum can carry hydrocarbon and the fluid waste up into the overlying formations, possibly contaminating water supplies To prevent any breakdown of the upper confining stratum during water flooding and waste disposal operation, an understanding of stress state in the confining layer and pore pressure distribution in the injection zone is required. The stress gradient in the underlying confining layer is usually, higher than in the injection zone, therefore it is less likely to break and has not been considered in the study. The pore pressure distribution in the injection zone was considered for steady state, transient, and pseudo steady-state flow systems. The radial and tangential stress profile In the confining layer were developed under complete wellbore stress relief and under the conditions of applied wellbore stress. Analytical equations for one dimensional models have been developed for the radius of potential fracture region in the confining layer considering an anisotropic stress field around the wellbore. Subsequently, the effect of pure pressure on the horizontal stresses In the confining layer has been studied to determine the potential fracture zones. Introduction Attempts are made to confine injection or disposal fluids in the desired injection zone for hydrocarbon conservation and ground water protection. For the confinement of disposal fluid the classical approach emphasizes on thewellbore Integrity and avoids interface with offset wells. Any fluid movement behind pipe is usually attributed to bad cementing operations and in attempt to mitigate this problem secondary cementing is usually recommended. The fluid movement behind the pipe could, however, be the result of local and regional stresses in the overlying formations. In the ground prior to injection, one of two possible conditions exists: there IS no pre-existing crack. a fracture starts from a pre-existing opening. To initiate a fracture in condition (I). the injection pressure should be higher than the tensile strength of the rock. However, in case the of condition (2). an injection pressure slightly higher than the least principal stresses would propagate the fracture if the crack is normal to the least principal stress 1 . When a fluid is injected into a permeable formation with no pre-existing crack. the pore pressure is increased near the wellbore and it will approach the far field pressure away from the well bore. If the injection pressure is lower than that required to reduce the compressive stresses across a plane ill the permeable formation to zero, then there can be no fracture in that formation. However high pore pressure can be transmitted into the impermeable confining stratum, if there exists a small defect in the form of a crack or notch at the interface between the confining layer and the permeable injection zone.

Abstract A procedure IS described that allows an operator to identify those beam-pumped wells which are operating at reduced efficiency. The logical sequence of steps to be followed in acquiring performance data such as power, dynamometer, fluid level, etc and the criteria to be used in determining the causes of inefficiency are presented. Following the logical methodology will reduce the time and effort required to perform the analysis. A field case is presented and discussed in detail. Introduction The need to increase oil production, reduce operating costs and increase net income from beam pumped wells, requiresan integrated analysis of the pumping system including the performance and interaction of all the elements: the reservoir, the wellbore, the downhole pump, the rod string the beam pump and the prime mover. Such system analysis can now be undertaken efficiently using portable lap-top based data acquisition systems in conjunction with appropriate transducers and a suite of analysis software 1 . Field experience undertaking such analysis in numerous wells has resulted in the development of a procedure which insures that good results are obtained with the minimum of effort. It is the objective of this paper to outline this procedure or methodology with the hope that it will be of use to production personnel throughout the industry. The end result of such system analysis should be the complete. Visualization of the performance of a given pumping well at a given time and a set of recommendations to be followed if significant improvements can be achieved. In general the following steps should be undertaken: Establish the well's inflow performance to determine if additional production is available. Determine the overall efficiency as a means to identify wells which are candidates for detailed analysis. Analyze performance of downhole pump and downhole gas separator. Analyze mechanical loading of rods and beam pump. Analyze performance of prime mover. Design possible modifications to improve existing system Implement changes and verify improvement. Although such procedure is greatly facilitated through the use of the Well Analyzer System 1 which includes all the necessary hardware and software components it can be undertaken with a number of other devices and components currently commercially available from other manufacturers. Methodology The steps to be followed in defining the performance of the system should result in the maximum of information with a minimum of time and effort. This is facilitated by having access La accurate and timely information about the well's characteristics, completion, well tests, etc. Although this may be initially time consuming it is generally a task that has to be completed only once. Some sort of data base management system is helpful in maintaining this information current. The information should be organized by well into a summary. Well Data File which can be accessed by most of the application programs that are likely to be used. This eliminates having to re-enter the data. Figure (l) shows an example of such Well Data File.

Abstract A new equation of sate-based technique for predicting minimum miscibilitypressures for CO 2 cured old systems has been developed. To verifythe technique, a total of 33 slim tube displacement tests were conducted.Reservoir fluids from four different pools were included in the testing. Slimtube displacement test results included measurement of oil recovery, pressuredrop, phase observations and effluent gas compositions, all of which wererequired to infer the experimental miscibility conditions. The application ofthe new technique successfully matched the experimental results over a widerange of reservoir temperature, 21.1 ° c to 78.3 °c, and fluid bubble pointpressures, 3490 kPA to 17330kPa. Introduction Within the literature there are many correlations (1–4) available to predict CO 2 miscibility pressure, However, these correlation do notadequately reflect oil composition and oil characterization. From previousexperience in hydrocarbon miscible solvent design (5), at any temperature thecontribution of each hydrocarbon component is different so hat solventmiscibility is a function of the oil composition. To overcome the limitationsof the current approaches, a new technique is proposed for predicting CO 2 miscibility pressureswhich uses a "tuned" equation of slate forprediction, The advantages of the technique are that cursory assessments onapplying CO 2 can be made inexpensively and the number of slim tubetests needed to verify prediction can be minimized, This paper describes thenew technique and documents all slim tube test data obtained to date such asoil recovery, pressure drop, phase observations and the effluent gascompositions required to experimentally determine miscibility conditions. Reservoir Fluids Table 1 gives the five oil compositions form the four pools examined in thisstudy. The specific gravity and molecular weight of the C 7 + fraction which were used in the equation of statecharacterization (6) are also shown in the table. The range of temperaturebeing tested is from 21 to 78 ° C and the bubble point pressure changes from3490 to 17,330 kPa for Oil-1 and Oil-4 crude respectively. New Prediction Technique The procedure used in this new technique is an extension of the pseudo ternarydiagram construction technique (7) for predicting miscible solvent gascompositions. That technique is based on blending tow hydrocarbon streams, drygas (DG) and LPG, at a specified design pressure and reservoir temperature todetermine the correct solvent composition. Since either pure CO 2 orCO 2 that contact impurities represents only a single identifiablestream, it is more difficult to identify separate stream compositions needed toconstruct the pseudo ternary diagram using an approach similar to that forhydrocarbon miscible processes. Based on an examination of the problem, arecommended approach for selection of the second stream needed for ternarydiagram construction is as follows: As a logical choice, the CO 2 -rich stream (Either pure CO 2 or CO 2 with impurities) is selected to be equivalentto a dry gas stream. This includes and CI and N2 impurities in theCO 2 stream. The second stream, which in the hydrocarbon miscible process is the LPGcomposition. Is made up of intermediates in the oil itself.

Abstract A new equation of sate-based technique for predicting minimum miscibility pressures for CO 2 cured old systems has been developed. To verify the technique, a total of 33 slim tube displacement tests were conducted. Reservoir fluids from four different pools were included in the testing. Slim tube displacement test results included measurement of oil recovery, pressure drop, phase observations and effluent gas compositions, all of which were required to infer the experimental miscibility conditions. The application of the new technique successfully matched the experimental results over a wide range of reservoir temperature, 21.1 ° c to 78.3 °c, and fluid bubble point pressures, 3490 kPA to 17330kPa. Introduction Within the literature there are many correlations (1–4) available to predict CO 2 miscibility pressure, However, these correlation do not adequately reflect oil composition and oil characterization. From previous experience in hydrocarbon miscible solvent design (5), at any temperature the contribution of each hydrocarbon component is different so hat solvent miscibility is a function of the oil composition. To overcome the limitations of the current approaches, a new technique is proposed for predicting CO 2 miscibility pressureswhich uses a "tuned" equation of slate for prediction, The advantages of the technique are that cursory assessments on applying CO 2 can be made inexpensively and the number of slim tube tests needed to verify prediction can be minimized, This paper describes the new technique and documents all slim tube test data obtained to date such as oil recovery, pressure drop, phase observations and the effluent gas compositions required to experimentally determine miscibility conditions. Reservoir Fluids Table 1 gives the five oil compositions form the four pools examined in this study. The specific gravity and molecular weight of the C 7 + fraction which were used in the equation of state characterization (6) are also shown in the table. The range of temperature being tested is from 21 to 78 ° C and the bubble point pressure changes from 3490 to 17,330 kPa for Oil-1 and Oil-4 crude respectively. New Prediction Technique The procedure used in this new technique is an extension of the pseudo ternary diagram construction technique (7) for predicting miscible solvent gas compositions. That technique is based on blending tow hydrocarbon streams, dry gas (DG) and LPG, at a specified design pressure and reservoir temperature to determine the correct solvent composition. Since either pure CO 2 or CO 2 that contact impurities represents only a single identifiable stream, it is more difficult to identify separate stream compositions needed to construct the pseudo ternary diagram using an approach similar to that for hydrocarbon miscible processes. Based on an examination of the problem, a recommended approach for selection of the second stream needed for ternary diagram construction is as follows: As a logical choice, the CO 2 -rich stream (Either pure CO 2 or CO 2 with impurities) is selected to be equivalent to a dry gas stream. This includes and CI and N2 impurities in the CO 2 stream. The second stream, which in the hydrocarbon miscible process is the LPG composition. Is made up of intermediates in the oil itself.

Abstract The displacement of oil by bottom water to a horizontal well in a verticallyconfined reservoir was studied experimentally in a Hele-Shaw cell. Thestability of the water-oil interface was investigated at different flow ratesand viscosity ratios_ A stable interface with higher recoveries was obtainedfor lower flow rates and for the cases where water and oil had the sameviscosities. The interface became unstable for higher flow rates and/orviscosity ratios. Fingering flow, which results ill bypassing of oil, reducesultimate oil recovery. For a given viscosity ratio, the higher flow ratesusually result in a lower oil recovery at breakthrough. However, in some caseswhere multiple fingers fanned, the oil recovery for higher flow rates was foundto be higher than that for lower rates. Gil recovery at breakthrough wascorrelated with a dimensionless flow rate. A novel numerical method was developed in this study to simulate [he pistontype of displacement found in Hele-Shaw cells. The advancement of the water-oilinterface and the corresponding oil recovery, as well as the pressurebehaviour, were predicted. A single finger was identified for different flowrates when the viscosity ratio was adverse. The simulation results were ingeneral agreement with corresponding experimental data.. Introduction Reservoirs underlain by bottom water zones are very common and constitute asignificant resource. The effective exploitation of these resources wouldprovide an offset to the oil production decline from lop quality reservoirs.Production from such reservoirs usually involves the completion of the wellnear the top of the reservoir so as to reduce the production of freewater. If the interfacial effect is ignored, then there are two kinds of force exertedon the interface during production: the viscous force that results from theflowing of fluids in the reservoir and the net gravitational force thatresults from the difference in density of two phases. For upwards displacementof oil by water, the gravitational force tends to stabl1iu: the water-oilinterface if water has the higher density than oil When the difference betweenvertical pressure gradients across the interface becomes greater than thedifference of gravity pressure gradients, the water-oil interface becomesunstable and viscous fingering occurs. Otherwise, the water-oil interface isstable. Figure 1 shows schematically the shape of a stable water cone for avertical well am.! a corresponding crest for a horizontal well. For verticallyconfined reservoirs, the cumulative oil production for isolated wells at thebreakthrough depends on the volume of water cone or crest underneath theproduction well (Butler, 1992). Cumulative oil production by bottom water drivecan be improved by horizontal wells in place of vertical wells because of theextended length of water crest (see Figure 1). When the viscosity of water is less than that of oil, which is the usual casein actual reservoirs, the upwards displacement of oil by bottom water may causethe interface lo become unstable at high production rate because of the adversemobility ratio between water and oil phases.

Abstract Large volume underground storage of hydrocarbons has been developed and practiced in Saskatchewan. Canada for many years. Use of these facilities has gained importance in recent years, due mainly to economical and environmental reasons. TransGas Limited currently has 42.8 billion cubic feet (Bcf) or 1206 million cubic meters (10 6 m 3 ) of working storage capacity. Of this total 14.1 Bcf (397 10 6 m 3 ) is held in solution mined bedded salt caverns and the remaining 28.7 Bcf(809 10 6 m 3 ) is held in depleted gas pools and gas caps. The total gas storage capacity, is 83.97 8cf(2366 10 6 m 3 ) of which 16.46 Bcf (463 10 6 m 3 ) is held in caverns and 67.51 Bcf (1902 10 6 m 3 ) is held in depleted pools. All of the depleted gas pools are located on the western side of Saskatchewan, mainly in Lower Cretaceous sandstone reservoirs. The washed salt caverns are scattered across Saskatchewan, all developed in the halite portion of the Middle Devonian Prairie Evaporite salt bed. Traditionally, TransGas Limited has developed its solution mined salt storage caverns from start to finish to supply Saskatchewan customers.. However, in the increasingly competitive North American natural gas storage market, TransGas Limited has had to look for innovative methods to develop cost effective storage facilities while providing new market development opportunities, One method of prodding natural gas storage service, used in two recent cases, is to convert existing salt caverns, previously used to Store Liquefied Petroleum Gas (LPG) products, to natural gas storage. TransGas Limited currently owns, leases and operates twenty-one (21) bedded sail caverns, of which six (6)caverns are currently LPG products, Three (3) additional bedded salt caverns are currently in the process of being Solution milled. This paper presents a general overview of TransGas Limited's present experiences with the conversion of LPG salt caverns to sweet natural gas storage service. Introduction Some of you today will have knowledge of the storage operations within Saskatchewan. TransGas pioneered the development of underground solution mined bedded salt caverns in Canada (and the United States as well) in the early 19605 at Melville, Saskatchewan. The sole purpose was for the dry or brine-free storage of natural gas. The equivalent caverns in the United States Were developed at Eminence, Mississippi in 1970. Prior to that time, salt caverns had been utilized by other companies for life storage of Liquefied Petroleum Gas (LPG) and other hydrocarbon liquid. Salt caverns permitted the highest rate of gas delivery; provided a secure and environmentally safe form of storage; permitted full product recovery; and required the least cost for construction; factors which still hold true today. In the 1990s, TransGas purchased outright or leased a total of six(6) former solution mined bedded salt, LPG caverns and converted them to natural gas storage. Three (3) of these caverns are located at Regina North and three(3) at Melville South (see Figure 1). One quarter of all TransGas cavern storage comes from converted LPG caverns. The pre-requisites for cavern conversions are identified in Figure 2.

Abstract Are all systems knowledge-based? What special characteristic(s) separate knowledge-based from non-knowledge-based systems. Our introductory remarks are aimed to stimulate discussion q{ what exactly constitute knowledge-based systems. We briefly explain and discuss three disparate knowledge based system applications under development in universities. One application falls under the heading of constraint oriented logic programming expert systems, the intelligent graphics interface (IGI) project. We describe System X, our natural language interface designed/or executive information access and manipulation as a knowledge-based system. We then explain how DB-Discover our system for discovering the implicit classifications and categorizations in a relational data base was designed and how it works. We conclude by discussing some ramifications of knowledge-based systems. Introduction Artificial Intelligence (AI) researchers noticed early on in the development of expert systems technology that expert systems derive their power from the knowledge they possess rather than on particular inference mechanisms or representational formalism (notwithstanding arguments of expressive power and notational efficacy). These later considerations, however, are certainly important for overall system performance and efficiency. This observation led to more robust problem solving and knowledge representation schemes and these, in turn, led to more efficient constraint propagation algorithms and effective inference implementations in modern approaches to building knowledge-based systems. At one time the terms expert system and knowledge-based system were synonymous; contemporary usage generally favours the more general and encompassing term, knowledge- based system. Architectural principles for knowledge based systems emerge when certain characteristics are identified. Knowledge is the central ingredient powering expert systems. Knowledge is often inexact and incomplete and often poorly specified. Novices become knowledgeable incrementally. Knowledge-based systems should be flexible as well. Many early AI systems simply did not qualify as knowledge- based systems even though their performance may have been admirable. For example, the general problem solver. To address these requirements, IGI has evolved the architecture shown in Figure 1. An IGI is a knowledge-based mediator between multiple real-time asynchronous control systems and multiple human operators working within an information bandwidth limited display environment. The mediation paradigm views IGI as an intelligent negotiator which uses various knowledge sources to manage the limited user interface resources. These limited resources include: host workstation(s) display capabilities existing interface protocol with the application database(s); and most importantly, the attention limitations of the human operator(s). Figure I illustrates this definition schematically. IGI uses a number of specialized knowledge bases to mediate the interaction of operators with the control system database. The mediation process is dialogue-based among operators, user interface objects and control system objects. The dialogues arise asynchronously by operator, program or application control. Dialogues internally establish a synchronous communication path between objects which may persist indefinitely. A dialogue consists of a sequence of one or more messages exchanged synchronously and bidirectional between objects. Figure 1 - Functional Overview of the lGI Architecture. (Available in full paper) Dialogues are persistent communication media required for IGI reasoning. We compare dialogues with the message pa

Abstract As part of a larger project, the relationship of horizontal well productivity and reserves to geology and reservoir properties has been examined at the Winter and Senlac heavy oil fields in Saskatchewan and at the Pembina field in Alberta. In the Winter Cummings Sand pool, vertical wells have not been successful in exploiting 36 million m 3 of heavy oil that occurs in a thick sand bar. Horizontal wells have average production of 22 m 3 /well/day and indicated ultimate reserves of 15.000 m 3 /well in the portion of the pool studied. Developed reserves are strongly dependant on standoff of the wellbore above the oil-water contact with at least 4.5 m of standoff required to produce beyond payout. In the Senlac Cummings-Dina pool, vertical wells are unsuccessful whereas horizontal wells are producing heavy oil at economic rates, The main sand body is possibly an estuarine deposit that occurs in a channel between continental and marine depositional realms. In the portion of the field studied, horizontal wells have ultimate reserves of 25,000 m 3 /well. Standoff of the wellbore above the oil-water contact IS not an obvious control of reserves developed but there is a positive correlation between reserves and wellbore length. Success of horizontal wells in the Pembina Cardrum light oil pool is strongly dependant on geology and reservoir properties. Horizontal wells located In areas of thick permeable conglomerate and permeable sandstone are successful whereas those drilled in low permeability reservoirs are failures. Introduction As part of a larger study, the authors have investigated the results of horizontal drilling in three sand pools in western Canada, The three pools are: Winter Cummings Sand Pool a producer of heavy oil in the L1oydminister area of western Saskatchewan Senlac Cummings-Dina Sand Pool which also produces heavy oil near L1oydminister and the Pembina Cardium Pool an important producer of light oil in western Alberta (Figure I). Winter and Senlac each have original oil-ill-place in occurs In thick sands having excellent reservoir properties. Both pools have extensive bottom water zones so that water coning has been a severe production problem in vertical wells because of the adverse mobility ratio of oil to water. Generally vertical wells have been uneconomic whereas there has been successful cold production from horizontal wells. The Pembina Cardium light-oil pool was discovered in 1953 and contains approximately 250 million m 3 (1.600 million STH) of recoverable reserves. The Cardium oil reservoirs are of interest for horizontal development because or their vast size and because there is a large amount of unrecovered oil-in-place despite secondary recovery being in place for a long period of time. At the time this project was undertaken. Amoco had drilled several horizontal wells in the study-area in the southwest part of Pembina field. Eight of these horizontal wells, drilled in areas of poor reservoir were unsuccessful whereas a horizontal well drilled in an area of good reservoir appears to be economic. This report is drawn from a larger study conducted under the sponsorship of Saskatchewan Research Council.

Abstract If the problem of fluid flow in a reservoir toward a well cannot beapproximated as a two-dimensional one, the analytical treatment becomes verydifficult. This is especially true when horizontal wells are involved and theend-effect cannot be neglected. There are, in literature mathematical expressions to calculate productivity ofhorizontal wells for various geometric configurations and many of these includethe effect of flow convergence. Some of them appear to be significantly inerror. The paper describes experiments with an electrolytic model. This approach usesthe analogy between the flow of fluid particles, driven by fluid flowpotential, in a porous medium and movement of ions, driven by electricalpotential, in an electrolyte. A simple and low-cost experimental setup permitsthe testing of various theoretical equations and the results arepresented. Introduction In many cases the flow of fluids in a homogeneous reservoir toward a well, vertical or horizontal, can be analyzed simply whenthe problem is two-dimensional. In some situations this approach may besufficient; but when, because of more complicated geometry, the effect ofconvergence toward the tip of a well (as, for example, in the case of avertical well partially penetrating a liquid-bearing porous matrix layer) hasto be taken into account, the difficulties of establishing the flow patternand, consequently, calculations of the well productivity, may besignificant. A common simplification is the assumption of a uniform flux along the activewell length. But this approach, as pointed out by Muskat 1 , does notgive the correct results, even for the relatively simple case of a partiallypenetrating vertical well and steady-state flow. The situation becomes more difficult for horizontal wells of finite lengthdraining bounded volumes. There is usually a problem of convergent flow towarda well in a plane perpendicular to the well's axis, but on top of this there isa three-dimensional convergence toward the tips of the well. Probably in somesituations when, for example, the well is "Very long and draining a thin layerand narrow pattern, this additional convergence can be neglected. But in anycase, an engineer should be at least aware of the order of error thisapproximation can create. There are several known formulae that are used for horizontal wellsproductivity predictions and which take into account various aspects of thegeometry of the problem, including well-end convergence. In some cases theformulae published by different authors differ significantly. Unfortunately, there is not much information available about the experimentalor field data regarding these problems. There is no= doubt that obtaining suchdata, especially in the field situation, may be difficult. It would requireflux measurements at various sections of a well, treated as isolated segments, but at the same time the flow inside the wellbore would have to be maintained.However, there is a possibility to model these conditions, at least forsteady-state flow in some situations, and to obtain experimental data. This canbe achieved by using the analogy between fluid flow in porous medium andelectric current flow in electrolyte.

Abstract Pipeline networks constitute the major bulk carriers for crude oil, natural gas, water and petroleum products in Western Canada. Each day millions of barrels ofcrude oil billions of cubic feet of natural gas as well as millions of gallons of water are transferred through pipeline networks from the sources to the users. Optimum operations scheduling of these pipeline. Systems can he used to increase the performance and reduce the energy consumption of pumping stations. A key factor in making such systems truly successful is the accuracy a/their demand prediction which is normally implemented with time series analysis. However, for many areas especially in Saskatchewan, weather, special events and other related parameters have major effects on the demand patterns which cannot simply be modeled with the time series techniques. This paper demonstrates how artificial neural networks (ANN) improve the demand prediction of pipeline networks. The water demand patterns were modeled by historical data and related variables using a single continuous perceptron (SCP) and multilayer perceptron (MLP). The implementation was based on real-world data from the City of Regina's pipeline networks. Weights of the SCP model were interpreted to determine relationships between demand patterns and related variables. Finally, comparisons between the models and performance improving techniques were discussed. Introduction The accuracy of demand prediction is an important parameter for the operation planning of pipeline network systems for crude oil natural gas and water transportation. For example the prediction of demand patterns can be used to find optimum pumping schedules. In the implementation of the intelligent system on monitoring and control of the water distribution system at the City of Regina, the knowledge of future demand has great effects on the operation performance of pipeline networks. Most of the existing demand predictors are based on time series analysis, especially the ARIMA (Autoregressive Integrated Moving Average) model. For example. Quevedo et al 1 used Box-Jenkins methodology to predict the water demand of Barcelona city's distribution network. However, it was found that the Box-]enkins's model is very sensitive to noise and is not appropriate for small data set. 2 In addition there arc a few other approaches such as were examined (The intervention components Ie were omitted in Equations 5.2 and 5.3 for clarity): Equation (5.1) (Available in full paper) Equation (5.2) (Available in full paper) Equation (5.3 (Available in full paper) Equation (5.4) (Available in full paper) Equation (5.5) (Available in full paper) Then all the components were combined to predict the water demand as shown below: Equation (5.6) (Available in full paper) For Equation 5.6, the prediction error was reduced to 3.5% of the mean demand value. However, the errors from the other pipeline networks were lying between 4.16% to 6.74%. A Disadvantage of this method is that model testing and modification is time consuming. By contrast, the ANN approach docs not involve human intervention transfer function analysis and model selection because it can automatically capture the model from appropriate training data: autonomous updating is also possible.

Abstract With renewed interest in enhanced oil recovery (OR) as existing fields become depleted, there is renewed interest in supplying CO 2 for flooding by CO 2 recovery from flue gas. This paper examines key design and operating issues and trends for new unit designs which hold the potential for reducing the costs associated with this source of CO 2 . Introduction In the mid-I980's, when crude oil prices were over $30 per barrel, several plants came on line in West Texas in the U.S. which used 30 wt% MEA with special inhibitors to extract CO 2 from boiler flue gas for use in OR. However, when the price of crude oil dropped dramatically, these plants quickly became uneconomic and were scrapped. The then existing pro=s licensors, Dow Chemical and Union Carbide, became less active in promoting these technologies and in 1989 Dow Chemical sold their FT-l process to Fluor Daniel Inc.. Fluor Daniel renamed the technology "ECONAMINE FG SM " and bas since licensed two commercial plants, one 320 te/d plant at Bellingham Massachusetts (U.S.A.) (1989) and also te/d plant for Sumitomo Chemical at Chiba, Japan (1993). Both plants sell CO 2 primarily to the food industry. The Bellingham plant, which bas a gas turbine exhaust flue gas source, was constructed by Fluor Daniel. The Sumitomo plant was constructed by Mitsubishi Heavy Industries (MHI) with the flue gas source from a boiler fired by a variety of fuels including heavy fuel oils. In addition, Fluor Daniel has completed a number ofstudies, several of which involved large-scale plants (1000 teld and larger). These studies were prompted by interest in using the CO 2 for OR and to avoid the new carbon tax being imposed in EC countries. In addition, Fluor Daniel designed two pilot plants for Japan which were used by the owners to study the process for greenhouse gas mitigation purposes. The results from the KEPCO pilot plant have been reported in the literature. 1,2 Much of the KEPCO work is applicable to OR applications as will be discussed later in this paper. Interest in greenhouse gas emission mitigation has spawned a number of processes for capturing .CO 2 from boiler stacks (hat differ considerably from capture by chemical solvents. None of these processes are presently offered commercially ant! most will remain in the laboratory study phase at least for the near future. The inherent cost of CO 2 capture and subsequent sequestration either in land formations or as clathrates in deep ocean trenches has proven to be expensive for any proposed process. Japan's Central Research Institute of the Electric Power Industry (Criepi, Tokyo) has concluded (hat the economics of CO 2 fixation and disposal will not be economical for fossil fuel plants and proposed other alternatives for limiting CO 2 emissions. 3 The Japanese government has co-funded a number of cooperative studies concerning CO 2 capture and sequestration. In Norway, which presently is imposing the highest carbon tax in the EC, studies are underway to develop novel CO 2 capture techniques (hat employ chemical solvents using membrane technology in the absorption step.

Abstract There are many factors which need to be considered when selecting a process for a given sweetening application. The system discussed in the paper is the application of chemical solvent processes to develop an economic way of reducing H 2 S levels in produced and fuel gas and the reduction of H 2 S levels from atmospheric tank systems. The chemical solvent processes are applicable at low pressure and relatively low temperature. Most commercially available chemicals for H 2 S sweetening have very similar operating principles. The efficiency of the chemical conversion equipment used has t0 be oriented towards the physical gas distribution and gas-liquid contact efficiency. Some key items to consider for the chemical reaction mechanisms in the solvent sweetening process are very essential for economic operation of the process. Most of the chemical sweetening produ.cts used are water soluble, therefore their efficiency is very low in dry gas production systems and thus the gas should be saturated Water based sweetening products have the potential to saturate lean gases to their natural saturation point. Foaming problems have resulted in serious deterioration of the chemical solvent efficiency. When bacteria development takes place in the produced water system, the bacteria growth is often inhibited by the use of biocides which contain aldehydes, The aldehydes forms the reaction base for the sweetening process, therefore their directed use either for bacteria or H 2 S sweetening is not obtained. Introduction Hydrogen sulfide can cause serious production operation problems due to corrosion, environmental, gas quality and safety concerns. The oil industry bas adapted two (2) methods for H 2 S removal. In large process systems, the use of Amine process is most common, in smaller processing systems, solvent extraction methods arc often used where a reaction product provides the opportunity for the H 2 S to ionize and attach to partially positive charged carbon atoms. When the H 2 S is generated by bacteria such as Sulfate Reducing Bacteria (SRB), the H 2 S level, can be reduced, disregarding the origin of the H 2 S, through H 2 S scavenging. This reaction products used for the H 2 S conversion all appear to have very fast reactions such as the glutaraldehyde and aerolein solutions. The reaction mechanisms and effects of salt and the organic phase to the reaction mechanism was studied by Elliot and Raymond. The following provides critical equipment design Parameters which can be utilized in the H 2 S scavenging, the advantages and drawbacks of some commercial chemical products are also outlined. Chemical reaction products for h 2 S sweetening The residual product developed in the H 2 S sweetening process is a very essential criteria in determining the utilization of the sweetening process for an in-line treating system. The chemical solvent efficiency and potential absorption by the sour gas arc the two other concerns associated with the sweetening process especially for in-line process development. Experimentation of several sweetening solvents have demonstrated drawbacks to its utilization. A detailed study was conducted by Fisher et al,. Saturated sour gas systems can increase CaCO scale deposition. The solid scale deposit creates an excellent environment for the culture growth of Sulfate Reducing Bacteria (SRB) which results in increased H 2 S levels down the line.

Abstract CL Consultants Limited has been involved in over seventy horizontal wells in S. E> Saskatchewan. From this experience has arisen a number of observations as to the tolls and skills needed in geosteering and the determination of reservoir quality. The mandate of the wellsite geologist is to place and maintain the wellbored in an predetermined location deemed as optimum. Samples, hydrocarbon detection, survery data, and LWD tools are constantly providing information to the wellsite geologist. How the geologist in the field uses this information to steer the bit to and within the reservoir and make determinations as to reservoir quality will be discussed. Introduction In 1989 CL Consultants was contracted to provide the Hydrocarbon Logging and the Wellsite geologist for a well in central Alberta. This project was to drill the Nisku horizontally. The depth of the zone was clearly outlined and a predetermined depth to go horizontally was defined Using the hydrocarbon logging equipment in particular the chromatograph, an oil contact was identified. Once this contact was reached it was clear what depth had to be maintained in the horizontal leg. The well proved to be very successful and upon completion of the job many observations were made and many questions where raised Correlation of the samples and the hydrocarbon data to the wireline data was very concise Porosity and reservoir quality could be defined while drilling. This, combined with the ability to steer the bit into the best reservoir made the proposition of horizontal drilling very exciting. Today, the skills tools and knowledge have greatly increased. With this expansion this techniques, the responsibilities of the well site geologist have changed and expanded. On a typical vertical well one geologist is responsible for tracking the wellbore as it drills through the various strata to a generally predetermined total depth Samples penetration rate and hydrocarbon detection provide many clues to that is occurring down hole. Once the primary objective is drilled the Geologist reports observations on porosity, fluorescence, and other hydrocarbon evidence The zone is then logged and all the information is compiled and the fate of the well is determined. Reservoir quality is based primarily on the open hole logs. The role of the geologist in a horizontal application creates an entirely different operational situation on the location. The entire process starts with the project geologist Interpreting the final is the most optimum location for the wellbore in the reservoir. Prior to drilling this "best spot" is outlined to the wellsite geologist in the form of a window that must be maintained. Once the wellsite geologist is on local on the build section has to be monitored and the correlation of the over lying strata must be carefully followed to ensure the reservoir is coming in according to the prognosis. Any changes in the build section are discussed with the project geologist to see if any angle changes to the build are required. Once the reservoir is encountered the window is maintained. If the rock changes a change in wellbore trajectory is necessary.

Abstract In heavy oil production by in-situ combustion, information on the relative importance a/the combustion and steamfronts is very useful in the development of good production strategies. To obtain this information/or two reservoirs, one containing heavy oil and the other a bitumen, a novel series of combustion lube tests was conducted. Contrary to conventional belief, the results showed that, before steam breakthrough, the produced oil properties were influenced much more by steam distillation than by cracking reactions. While the combustion front mobilized substantial amounts of oil, almost all a/it remained behind the steam front. This oil displayed huge differences in properties and composition compared with the original oil. Overall, the results helped to provide some guidelines for developing improved strategies for fireflood operation, and emphasize some requirements for the prediction of fiireflood behaviour. Introduction As a thermal method of improved oil recovery, in-situ combustion has the advantage of delivering heat to a formation very efficiently. Consequently, the prospect of using in-situ combustion to produce heavy oils has continually attracted interest, particularly for thin formations where heat losses to non-productive horizons arc large. The results of field trials l-3 in such reservoirs have varied widely, suggesting that the method may be strongly affected by the approaches used to apply it. For the development of efficient strategies for in-situ combustion, it is important for engineers to understand its many mechanisms and their impact on oil recovery. Islam et al 4 have demonstrated the importance of the gas flooding effect on oil recoveries during in-situ combustion. Other phenomena such as the combustion and steam fronts, must also affect oil production. During dry combustion, the steam zone may be small, particularly in thin reservoirs. In this case, the combustion and steam fronts act almost as a unit. During wet combustion, however, when water is injected behind the combustion front to scavenge heat and deliver it in the form of steam, to the oil-bearing zone, the steam front can advance well ahead of the combustion front. In this case, several important questions arise. Does the steam front dominate the oil displacement process, leaving the combustion front to function only as the heat source that sustains the steam front, or does the combustion front also serve to displace oil? To what extent does each front affect the properties of the displaced oil? In an effort to, answer these questions, a series of four combustion tube tests was carried out. These tests were designed to show the effects of each front in two reservoirs, one representative of a heavy oil, and one of a bitumen. Equipment The experiments were conducted in a heavy-walled Incolloy 800 combustion tube with an inside diameter of 5.37 cm, a total inside length of 162.3 cm, and a wall thickness of 0.49 cm. No pressure jacket was employed. The maximum working pressure was 4.2 MPa (610 psig). Pack temperatures at the centreline were measured with 25 thermocouples inserted through fittings welded in a helical pattern on the tube wall.