We study three-phase flow in water-wet, oil-wet, mixed-wet, and fractionally wet sandpacks. We make the mixed-wet pack by invading a water-filled water-wet pack with crude oil and aging it for a week. This process mimics wettability changes in reservoir settings, leading to a realistic arrangement of wettability at the pore scale. We characterize the wettability of each sand pack by measuring the capillary pressure curves. We obtain the oil and water relative permeabilites during three-phase gravity drainage, by measuring the saturation in situ using computerized tomography scanning. In an analog experiment, we measure pressure gradients in the gas phase to obtain the gas relative permeability. Thus we determine all three relative permeabilities as a function of saturation for each wettability. We find that under uniform wetting, the relative permeabilities of the most-wetting phase ͑water in a water-wet pack, oil in an oil-wet pack͒ are similar. However, the relative permeabilities of the intermediate-wet phase ͑oil in a water-wet pack, water in a oil-wet pack͒ are very different at low saturations, with spreading oils showing a characteristic layer drainage regime. The mixed-wet pack also shows the layer drainage regime. We also find that the gas relative permeability is smaller in an oil-wet medium than in a water-wet medium. We explain the observations in terms of wetting, spreading, and the pore scale configurations of fluid. Materials and MethodsWe chose sandpacks as our porous media because they are easy to characterize and they can be easily sectioned for destructive saturation measurements. We used clean industrial sand ͑No. 60, Corona Industrial Sand Co., Corona, CA͒ which was initially waterwet. The sand was put through a size 120 sieve to remove any fine particles. We made 15 kg of the sand oil-wet by soaking initially dry sand in a mixture of 20% crude oil ͑Thums Inc., Long Beach, CA͒ and 80% iso-octane for 24 hours. 14 This oil-wet sand was then rinsed with iso-octane and air dried. The fractionally wet sand was a 50-50 mixture of the oil and water-wet sands.Mixed-wet sandpacks were created from water-wet packs, first by saturating the pack with a 0.01 M NaBr brine ͑pH 4͒. 14 We then displaced the brine with five pore volumes of the crude/isooctane mixture, which yielded a pack with a water saturation of S w Ϸ0.2. The column was then left to age for a week. After the aging the crude mixture was displaced with iso-octane until the

TX 75083-3836, U. S. A., fax et-972-952.9435. AbstractWe measured three phase relative permeabilities for gravity drainage using a dual-energy medical CT scanner modified to scan vertical cores, Independent measurements of two saturations as a function of time and distance along the length of the core were made from which relative permeabilities were found. Three phase (air/oil/water) gravity drainage experiments were performed on systems with different spreading coefficients and at different initial conditions. Experiments were run on both consolidated and unconsolidated porous media. The results were compared to measurements of three phase flow in capillary tubes, micromodels and to predictions from network modeling.We find that at low oil saturation k,. N S; for hexane and octane as the oil phase. This functional form of relative permeability is consistent with the drainage of oil layers, wedged between the water and gas in crevices of the pore space. For decane, which is non-spreading, the layer drainage regime was not observed. At higher oil saturations kro -+ S: with a x 4 for spreading and non-spreading systems. Within the scatter of the experimental data, oil and water relative permeability are functions only of their own saturations and independent of initial conditions.

The most widely used method of thermal oil recovery is by injecting steam into the reservoir. A well-designed steam injection project is very efficient in recovering oil, however its applicability is limited in many situations. Simulation studies and field experience has shown that for low injectivity reservoirs, small thickness of the oil-bearing zone, and reservoir heterogeneity limits the performance of steam injection. This paper discusses alternative methods of transferring heat to heavy oil reservoirs, based on electromagnetic energy. We present a detailed analysis of low frequency electric resistive (ohmic) heating and higher frequency electromagnetic heating (radio and microwave frequency).We show the applicability of electromagnetic heating in two example reservoirs. The first reservoir model has thin sand zones separated by impermeable shale layers, and very viscous oil. We model preheating the reservoir with low frequency current using two horizontal electrodes, before injecting steam. The second reservoir model has very low permeability and moderately viscous oil. In this case we use a high frequency microwave antenna located near the producing well as the heat source. Simulation results presented in this paper show that in some cases, electromagnetic heating may be a good alternative to steam injection or maybe used in combination with steam to improve heavy oil production. We identify the parameters which are critical in electromagnetic heating. We also discuss past field applications of electromagnetic heating including technical challenges and limitations.

We provide a general framework for interpreting heavy oil solution gas drive experiments. Evolution of the gas phase below the thermodynamic bubble point is investigated using dimensionless scaling groups and mechanistic modeling. The role of initial solution gas oil ratio, oil viscosity and depletion rate on the early growth of the gas phase is discussed. Factors leading to the coalescence of gas bubble clusters and development of bulk gas flow are illustrated with examples, along with an empirical correlation for critical gas saturation (Sgc). Two carefully planned and monitored depletion experiments are conducted to evaluate the pre and post-Sgc behavior. These experiments are CT-scanned to obtain in-situ saturations, that are used in conjunction with differential pressure measurements along the core to obtain phase mobilities. Implications of conducting experiments at pressure depletion rates and gradients substantially higher than field rates are discussed. Introduction and Study Objectives It is widely reported that many heavy oil reservoirs have shown higher oil production and recoveries than would be normally expected from conventional reservoir engineering principles [1]. While sand co-production and reservoir compaction may contribute to the increased production rates in many cases, the role of solution gas drive mechanism is of great significance. Several experimental [2–11] and theoretical [12–19] studies have been conducted to understand this phenomena. Some of the key findings are summarized in Appendix A and B, respectively. While there are some good summary and critical review papers [1,20] on what has been done, to the best of the authors' knowledge there is no documented work that provides an overall framework for interpreting heavy oil solution gas drive experiments and relates them to field depletion rates. The key objectives of the current paper are as follows:Use mechanistic modeling and dimensionless scaling groups to relate pore level mechanisms of bubble nucleation and growth to macroscopic observations of heavy oil solution gas drive.Perform carefully planned and monitored depletion experiments under reservoir conditions, having the following salient features:–CT scanning for in-situ saturation measurement–Overburden pressure to maintain suitable stress–An accurate collection system, especially for gas production.Interpret pre and post-Sgc behavior of experiments conducted and put it in context with past experimental studies.Discuss implications of conducting experiments at pressure depletion rates and gradients substantially different from those observed in the field. Framework for Interpreting Heavy Oil Solution Gas Drive Experiments The mechanism of gas evolution during primary depletion maybe broken into three elements:Bubble nucleation and growthBubble coalescence and start of bulk gas flow (at the critical gas saturation)Two-phase flow of gas and oil. Note the gas may be dispersed under certain conditions.