Summary Flow pattern, pressure drop, and water holdup were measured for oil/water flow in horizontal, hilly terrain (±0.5 and ±3°), and vertical pipelines at a temperature of approximately 35 (± 5) °C and a pressure of approximately 245 kPa using the large-scale multiphase-flow test facility of Japan Oil, Gas and Metals Natl. Corp. (JOGMEC). Test lines of 4.19-in. (106.4-mm) inner diameter (ID) and 120-m total length were used, which included a 40-m horizontal or hilly terrain (near-horizontal) and a 10-m vertical test section sequentially connected. The flow pattern was determined by visual observation with video recordings, and a flow-pattern map was made for each condition. New flow patterns were identified for horizontal and hilly terrain flow, such as oil flow in a snake-like shape at top of pipe at high rate of water flow, and water flow at bottom of pipe at high rate of oil flow. New holdup and pressure-drop data are presented for each flow condition. Flow rate and inclination angle influence holdup and pressure-drop behaviors. In vertical flow, when the oil superficial velocity exceeds a certain value, the pressure drop decreases exponentially as the superficial oil velocity, vSO, increases. Slippage between the phases was analyzed using the measured water holdup plotted against the input water cut with inlet-oil flow rate as a parameter and slip velocity vs. measured water holdup. It was found that the slippage changed significantly with slight changes in inclination angle. This paper provides new experimental data of flow pattern, water holdup, and pressure drop measured particularly at horizontal, hilly terrain, and vertical conditions, with large-diameter pipes. This is indispensable information for developing reliable prediction models for oil/water two-phase and gas/oil/water three-phase flow in pipelines. Introduction In the petroleum industry, the joint flow of two immiscible liquids such as oil and water in pipes commonly occurs at facilities for production and transportation of oil (i.e., horizontal, inclined, or vertical pipes) in wellbores and flowlines. In offshore fields, these pipelines can be of considerable length before reaching the separator facilities. The pressure required to transport the fluid over long distances is highly influenced by the pressure drop that can be significantly affected by the mixture properties of the oil and free water. As the amount of free water increases as the field matures, a reliable prediction of pressure drop and water holdup is extremely important for the optimum design of pipeline systems in the industry. For a two-phase mixture of oil and water flowing together in a pipe, different internal flow geometries or structures can occur, depending on the flow rates of the two phases and the geometrical variables of the pipes, as well as the flow conditions and physical properties of the phases. The different interfacial structures are called flow patterns. Knowledge of the flow patterns that could occur under a given set of conditions leads to better prediction of oil-/water-flow behavior. In addition, accurate interpretation of experimental data requires reliable prediction of flow pattern, water holdup, and pressure drop. The flow characteristics of oil/water mixtures are generally different from gas/liquid systems. In oil/water flow, the different flow structure is mainly caused by the small buoyancy effect and lower free energy at the interface, allowing the formation of shorter interfacial waves and small dispersed-phase droplet size. Therefore, the results of gas/liquid flow cannot be applied directly to oil/water flow in most cases.

The experimental study includes phase behavior tests, measurements of interfacial tension (IFT), and coreflood tests. Effects of alkali (sodium carbonate) additives on micro-emulsion phase behavior were determined for a reservoir oil. No emulsion phase was generated for low surfactant concentrations without alkali additives, while an emulsion phase was always formed with alkali. Measurements of the dynamic IFT revealed that when the alkali concentration is as high as 0.7–1.0 wt%, the IFT rapidly increases after about 100 minutes, which suggests that the system shifts from the middle to upper-phase equilibrium. Based on tertiary alkali-surfactant-polymer (ASP) coreflood tests with surfactant and polymer concentrations of 1.0 wt% and 1200 ppm respectively, an optimum alkali concentration was confirmed as 0.2 wt% that allowed the highest recovery. Using the optimum ASP concentrations, another series of corefloods were conducted to evaluate the effect of remaining oil saturation at the start of ASP flood. Mobile oil helps alkali and surfactant slugs form an oil bank. In order to examine the oil recovery by high alkali and low surfactant concentrations (1.3 and 0.1 wt%, respectively), corefloods were conducted by changing the injection volume and scheme. Only a small incremental recovery is obtained with ASP slugs higher than 0.2 PV each. The injection scheme of polymer slugs is confirmed to be crucial for ASP flooding. Parallel coreflood tests as physical simulation of ASP flood in a two-layer system showed recovery performances complicated not only by permeability contrast but also by small-scale heterogeneity within inidividual cores. Acceptable matches between coreflood experiments and one-dimensional flow simulation were obtained by modeling ASP flood with recovery mechanisms of the surfactant-polymer flooding.

This study concerns nonisothermal single- and two-phase flow of a single-component fluid (water) in consolidated porous media. Linear flow experiments through cylindrical consolidated cores were performed. Both natural (Berea) and synthetic cement-consolidated performed. Both natural (Berea) and synthetic cement-consolidated sand cores were used. Fabrication of the synthetic sandstones was important to permit reproducible fabrication of high-porosity, low-permeability sandstones with thermowells, pressure ports, and glass-tube capacitance probe guides cast in place. Both hot-fluid and cold-water injection experiments were carried out in natural and synthetic sandstones. The thermal efficiency of hot-water and cold-water injection was found to depend on heat injection rate: the higher the heat injection rate, the higher the thermal efficiency. One important result of this study is that much of the previous work with nonisothermal single-phase flow in unconsolidated sands may be extended to consolidated sandstones despite the differences in the isothermal flow characteristics of these systems. In two-phase boiling flow experiments, hot, compressed liquid water entered the upstream end of the core, moved downstream, started vaporizing, and flowed through the remainder of the core as a mixture of steam and liquid water. Significant decreases in both temperature and pressure occurred within the two-phase region. Even for large temperature changes, it was found that two-phase flow can be nearly isenthalpic and steady state if heat transfer between the core and the surroundings is at a low level. Introduction Geothermal energy is being given much attention as a new source of energy. Prime questions in geothermal energy extraction are (1) how much energy can be recovered, and (2) how fast can it be extracted? To find useful answers to these questions, the basic nature of the boiling flow of water in porous media must be understood. Literature on oil recovery by hot-fluid injection and underground combustion presents some of the important features of nonisothermal, two-phase flow that appear pertinent to geothermal reservoirs. The injection of hot water to effect oil recovery was commonly considered before 1930. In 1930, Barb and Shelley mentioned a rumor that hot-water flooding had been tried in New York State and abandoned because of excessive cost. The heating and economic results of hot-water injection were evaluated in this pioneering study. pioneering study. The next study of heat transport in a formation caused by hot-fluid injection was presented by Stovall in 1934. Both laboratory and field experiments were described. Field determination of both wellbore heat losses and vertical losses from a heated formation were described in this remarkable study. Apparently, the next study of vertical heat loss on hot-fluid injection was published by Lauwerier in 1955. It was assumed that injection rate, Vw, and temperature, Ti, would remain constant; thermal conductivity in the direction of flow was zero; and the thermal conductivity in the flooded layer perpendicular to the direction of flow was infinite so that the temperature in the flooded layer, T1, was always constant at a given location in the flooded zone. Prats has called the latter condition the "Lauwerier assumption." The conductivity in the overburden and underburden, 2, was assumed to be finite and constant. The loss of heat from the injected fluid to the adjacent strata resulted in a decrease in temperature in the direction of flow. Lauwerier derived the temperature both in the injection interval and the adjacent strata as a function of time and distance. In 1959, Marx and Langenheim presented a solution for a heat-loss problem related to the one considered by Lauwerier, but where the heated region remained at a constant temperature equal to the injection temperature. Vertical heat loss reduced the size of the heated region. SPEJ P. 137

Experimental and model studies were performed on two-phase flow behavior at high-pressure conditions. The experiments were conducted using nitrogen and water in a test loop of 106.4 mm diameter pipe with inclination angles of 0°, 1°, and 3°at 2060 kPa. The liquid holdup data of 81 runs for each inclination angle were analyzed to identify the flow pattern.The mechanistic model developed for low pressures was modified for high-pressure conditions. The model first detects the flow pattern, and then calculates liquid holdup and pressure drop based on the flow pattern. For dispersed-bubble flow, the critical bubble size mechanisms were also applicable at high pressures to predict a flow region in the flow pattern map, and the slip model of liquid holdup showed better matches with the experimental data than the non-slip model. For stratified flow, the flow region in the flow pattern map extended to higher liquid flow rates than at low pressures. Sequential application of the Taitel _ Dukler and Bendiksen _ Espedal criteria could correctly identify the stratified and non-stratified flow transition, and the Lockhart _ Martinelli correlation based on the shear stresses could evaluate the liquid holdup much better than the common correlation based on the material balance. Elongated-bubble flow changed directly into dispersed-bubble flow as the liquid flow rate increases. Excellent performance of the model was demonstrated by error analyses of liquid holdup and pressure drop calculations.

Fracture-induced anisotropy can lead to observable azimuthal variations of seismic attributes that then can be used for characterizing a fracture system. Unfortunately, abnormal transmission losses along raypaths also can result in similar azimuthal variations leading to uncertainty in such fracture determination. Using a physical model containing gas-filled fractures, we investigate the impact of abnormal transmission loss on fracture detection from ultrasonic data in a laboratory setting. Recorded reflection amplitudes and traveltimes are used to study ultrasonic responses to the presence of the gas-filled fractures and to understand observed azimuthal attribute anomalies. Experimental results from this study highlight the pitfalls in using azimuthal attribute variations as indicators of the presence of fractures when abnormal transmission attenuation is significant.