SPE 14223Summary. Solvents and light' ends of cmdes ze frequently used as difuen& to facilitate pumping and pipeline kmspotion of heavy cmdes. The use of solvent done for in-situ recovery of heavy oil tends to be limited because of its high cost; however, the usc of solvent as an additive to steam processes has been tested both in the laboratory and in the field. The results of these tests are mixed. fn this study, we use numerical experiments to defineate the recovery mdanism of a steam-slug process when solvents are present. A gocd understanding of the mechanism will help provide an interpretation of the conditions under which solvents can improve steam oil recovery. The study focuses on the use of small quantities of solvent-i. e., no more than 10% of the steam volume. IntroduatlorrSteam injection is widely used for heavy-oil recovery. Using small amounts of ,solvent (fess than 10% of steam volume) with steam has pOtentiaJ io increase oil recove~. While solvent is known to benefit oif mcove~by reducing oil viscosity, the use of small amounts is insufficient to affect a significant volume of oif. In this simulation study, we found that a smafl amount of solvent can im. prove sweep by creating a mobility transition zone. When coinjected with steam, the vaporized solvent travels with steam and condenses in the cooler regions of the reservoir. R then mixes with oil, creating a transition zone of Iower-vismsity fluid between the steam and the heavy oif. The mobiMy ratio of the displacing and displaced fluids is improved, thus suppressing viscous fiigering and improti!ng sweep.The placement of the solvent in &e resemoir is crucial to the process performance because it determines the location of the hansition zone. The solvent placement is controlled by the injfxtion procedure and the solvent volatihiy. For example, when solvent is injected before steam, or a heavy soIvent witi low voladlity is coinjected with steam, incremental recovery is poor because of inadequate transition-zone generation in the reservoir.Although deftition of solvent volafll~depends on reservoir pressure and temperamre, fight solvmts generally include C02, ethane, propane, amd other gases, while heavy solvents generally include hydrocarbon liquids in the C 16to Cm range. Naphtha will be considered a medium solvent. Light solvents give earlier recovery and~eater recovery efficiency in km of less solventloss. Medium solvents give tie most improvement @ total oif production at somewhat higher solvent loss. Heavy solvents do not improve recovery. Previous WorkThe useof various solvents in conjunction with steam has been reporkd in the literature. Balier, Farouq .4di and Snyderl studied naphtha injection before steam injection in a two-dimensional (2D) vertical mcdcl Elfed with tar sand. Naphtlm was found to be highly effective in opening a steam flow path in a homogeneous sandpack but broke through quickIy in a pack with a highly permeable channel, AIikban and Famuq .41i2 further studied the steamdrive solventslug process in a linear celf model pac...

No abstract

CoF@ght 1993, Society ofPe!mlawn Engineers, Inc. This psfxir was pfapamd W Prasantarcm at tha 199S SPE Gas Technology Sympsium held m Calgary, Aibmfa, Canada, 1S-18 March 199S This pspr was selected for~eaentatii by an SPE Program Commrttee following revm'w of informatii corstamad in an abstrsd sulxsitbsd by the authcf( s). Contents of the paper, as pm?a?mted,have not k raviewed by ttw Society of Petroleum Engineers and are subjecl to wrac.tii by the auttvx(s). W material, as presented, does not necessarily reflect any postron of ttsa Scrcisty of Petroleum Enginaera, its offsxrs, or members Papws presented at SPE meetings am aubjad to publiitiem review by Editmial Committees of the Society of Petroleum Engirwars. Elec@xic rejxodudkm, distriiticm, or storage of any pafl of this papsr for cemmercts( purpses Wfscwt tha '#M& consent cd the SocWy of Petroleum Enginaars is pd_ibited Permiss"km to mpmduce h print is restricted to an ebstract of not more than X0 W* illustraticms may not be WPW, The abstract must contain conspicuous adaxnvladgment c+ wtt8ra and by wfvxn the ps$ar was presented. Write Librarian, SPE, PO. E&s SS3S%, R&srdscm, TX 7S9S2-SS26, U.S.A., 1SS01 -972-SS2-9435. AbstractNcn-Darcy flow increases the pressure drop required to establish a desired gas well production rate, thus decreasing productivity. This increased pressure drop is exacerbated by the liquid drop-out and build-up which occurs in gas condensate wells. The additional pressure drop caused by the tw~phase non-Darcy flow can have a dramatic effect on the flowing bottom hole pressure required to maintain producing rates, especially in high rate gas condensate systems. Neglecting or underestimating this effect will cause optimistic predictions of the maintenance of gas rate plateau.Literature values for measured inertial coefficients, (which quanti& non-Darcy flow pressure drops) show a scatter of a one to two orders magnitude for a given permeability. Measurements of inertial coefficients (betas) may or may not agz"ee with literature reported values.Use of literature reported permeability -beta relationships to calculate beta may not be appropriate and may lead to significant errors. These errors increase with increasing flow rate and decreasing permeability.This paper presents transient flow beta measurements obtained in dry core samples and in samples containing different saturations of immobile liquid. Results from core samples from three different reservoirs and one outcrop are included. A method has been developed to estimate the effective beta at different liquid saturations using core samples containing various saturations of solidified paraffin wax that mimic an immobile condensate phase. Continued validation of this technique will allow simplified and inexpensive beta measurements as fimctions of saturation and permeability A relationship has been derived which allows the inertial coefficient to be estimated as a tlmction of effective permeability and effective porosity. This relationship appears to hold for almost all of the core sa...

Summary This paper describes a simulation study on the effect of visbreaking on heavy oil recovery during steam injection processes. The kinetic rate constant for in-situ visbreaking was derived from in-house kinetic data and then used in conjunction with a thermal compositional simulator to assess the effect of visbreaking on recovery through a series of numerical experiments. Various steam injection strategies were tested and the effect of visbreaking studied. In some cases, physical heating, not thermal visbreaking, was the dominant recovery mechanism. In other cases, visbreaking had a large effect on recovery. The difference was primarily a result of the placement of the visbroken oil with respect to primarily a result of the placement of the visbroken oil with respect to the direction of flow. In cases where the visbroken oil zone was perpendicular to the flow, it formed a mobility transition zone that perpendicular to the flow, it formed a mobility transition zone that improved sweep, thus enhancing oil recovery. Introduction A major constraint in the production of heavy oil and tar sand bitumens is the high viscosity. Thermal recovery techniques, such as steam flooding, are effective in temporarily lowering the oil viscosity and enhancing its recovery. Field evidence, however, indicates that a permanent reduction in oil viscosity, or visbreaking, often accompanies these processes owing to the cracking of the oil. In one incident, oil as light as 21 deg. API [0.93 g/cm ] was produced from steam stimulation in a reservoir produced from steam stimulation in a reservoir containing 11 deg. API [0. 99-g/cm 3 ] heavy oil. Steam distillation could not account for this phenomenon because the oil had a very small fraction that boiled below the steam temperature. In the laboratory, numerous experiments confirmed that heavy oils experience significant visbreaking at temperatures in the range of 500 to 700 deg.F [260 to 371 deg. C). Thermal visbreaking, a well-known refinery process for upgrading heavy crudes, has been subjected to much study. Few studies, however, were conducted under the relatively mild conditions commonly encountered in insitu steam recoveries. In-situ visbreaking is characterized by mild decomposition, minimum coke formation, and the retention of the products in the liquid phase. Heavy components, such as asphaltenes, are cracked into lighter components that act as an internal solvent, thereby lowering oil viscosity. Little is known of the importance of in-situ visbreaking on heavy oil recovery. To the best of our knowledge, no studies on the subject have been reported in the literature. Shu and Venkatesan reported a kinetic study of in-situ visbreaking. Laboratory visbreaking data for a Cold Lake oil were collected at 500 to 617 deg.F [260 to 325 deg. C] with residence times of up to I month. An algorithm was derived that allowed the calculation of a kinetic rate constant directly from viscosity data. The rate constant was used in the present study as input to the simulator. A compositional thermal process simulator with capability to handle reaction kinetics was used. We investigated the effects of visbreaking in three modes of steam recovery: cyclic steam stimulation, continuous steamflooding, and a steam slug process. Kinetic Considerations it is instructive to first consider some kinetic aspects of visbreaking and to delineate the rate-influencing parameters. The visbreaking reaction may be represented by the parameters. The visbreaking reaction may be represented by the following first-order reaction. (1) The extent of visbreaking can be represented by the conversion of the heavy oil. (2) or (3) where CA is the concentration of Component A, and k is represented by the usual Arrhenius expression: (4) For the Cold Lake oil under study, w= 2.952 x 10 day -1 and E= 31,800 cal/g mol, as determined in the earlier study. if we substitute Eq. 4 into Eq. 3, we obtain (5) P. 474

No abstract