Foam Propagation on Semi-Reservoir Scale
- Frode Vassenden (SINTEF Petroleum Research) | Torleif Holt (SINTEF Petroleum Research) | Amir Ghaderi (SINTEF Petroleum Research) | Arild Solheim (IPRES)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- October 1999
- Document Type
- Journal Paper
- 436 - 441
- 1999. Society of Petroleum Engineers
- 5.5 Reservoir Simulation, 5.4.1 Waterflooding, 4.1.2 Separation and Treating, 5.4 Enhanced Recovery, 5.3.4 Reduction of Residual Oil Saturation, 5.7.2 Recovery Factors, 4.3.4 Scale, 4.2.3 Materials and Corrosion, 1.6.9 Coring, Fishing, 2.4.3 Sand/Solids Control, 5.5.8 History Matching, 1.2.3 Rock properties, 2.5.2 Fracturing Materials (Fluids, Proppant), 5.3.2 Multiphase Flow, 5.2.1 Phase Behavior and PVT Measurements, 4.1.5 Processing Equipment
- 0 in the last 30 days
- 348 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
Foam propagation in co-injection of gas and surfactant solution has been studied in a 10-m-long flow apparatus, equipped with pressure ports and fluid sampling valves for every 1 m. The data have been compared to the results of a core scale foam flooding experiment with the same porous medium.
The propagation experiments on the 10 m scale revealed that the foam front propagated significantly slower than the injected fluid front. It appeared that foam propagation was not limited by surfactant transport, but was delayed due to the presence of oil in the porous medium. The experiments have been interpreted with the aid of a numerical foam simulator.
Oil is often effectively displaced by gas. Due to the low density and low viscosity of gas, it may be difficult to achieve a good macroscopic sweep efficiency, however. It has been found that the use of foam may reduce gas mobility, and thereby improve the sweep of gas. Foam can be formed within the reservoir when gas and surfactant solution flow together.
A successful well treatment with foam is critically dependent on placement of foam to the desired depth in the reservoir. The injection time required to reach a given depth depends on the propagation velocity of the foam. For the design of a foam treatment, prediction of foam propagation becomes an important issue. The aim of the present study was to clarify which mechanisms determine the foam propagation velocity, and to find out how foam propagation on the reservoir scale relates to foam data obtained in conventional coreflood experiments.
The literature provides some observations of foam propagation rates at various conditions. At the Kern River steam foam pilot,1 it was found that the formation of one volume of a C16/18 ?-olefin sulphonate (AOS) foam within the formation required injection of 1.5 volumes of surfactant solution and 700 volumes of steam at reservoir conditions. This propagation delay was interpreted as mainly caused by surfactant retention, but it was also suggested that inefficiencies in foam generation and bubble transport could have further slowed down the growth of the foam zone.
Also, Irani and Solomon2 observed slow foam propagation, in slim-tube studies of AOS-stabilized CO2 foams. In experiments without surfactant in the slim tube before foam injection, the foam propagated significantly slower than the injected fluid fronts. In experiments with gas slug injection into porous medium presaturated with surfactant solution, piston-like propagation at the velocity of the injected gas front was observed. When oil was present in the porous medium, there was significant propagation delay observed also during gas slug injection, however.
In co-injection experiments in cores with the surfactant adsorption already satisfied, Kovscek et al.3 observed piston-like foam propagation with no retardation of a C16/18 AOS foam. Similar observations were made by Osterloh and Jante.4 Kovscek et al. also performed experiments where the core was free from surfactant at the start of foam injection. Then, retardation of the foam front was found, however, demonstrating how surfactant adsorption retards foam propagation.
Aarra et al.5 found significant retardation in experiments in cores with residual oil saturation after gas flooding. The cores were presaturated with C14/16 AOS surfactant solution. At a total injection rate of gas and surfactant solution of about 3 m/d (interstitial velocity), the foam propagated with a velocity of 0.036 m/d. In this case, surfactant adsorption cannot explain the slow propagation, and the oil remains the most probable cause for the retardation.
Mannhardt and Svorstøl6 also reported slow propagation in systems with surfactant adsorption satisfied. At an injection rate of 1 m/d, the foam used 10 days to propagate the first 40 cm of a core which had 19% pore volume (PV) oil saturation. Without oil, the foam traveled 40 cm in less than half a day, which corresponds approximately to the velocity of the injected fluids. This also points to the oil as one source of propagation delay. At the employed conditions, the foam strength was found to be sensitive to the presence of oil.
The literature on oil-free experiments with adsorption satisfied demonstrates that the propagation delay is not intrinsic to foam per se. A quantitative understanding of all effects that control the propagation delay is lacking, however. The present study aims at improving this understanding. The approach taken has been to study foam propagation in well-characterized systems, over large distances (10 m). The porous medium has been characterized by laboratory experiments on the usual core scale, with respect to relative permeability, capillary pressure, and foam flow and surfactant adsorption properties. Then, foam propagation was studied on the semi-reservoir scale, and compared to modeling based on core scale data, in order to learn how to use core data to predict foam propagation in the reservoir.
The semi-reservoir scale flow experiments were carried out at reservoir conditions in a 10-m-long sandpack. The container for the 10 m sandpack was a specially constructed assembly of ten 1 m long tubes made of the corrosion resistant alloy Hastelloy C-276. The outer and inner diameters of the tubes were 25.4 and 18.1 mm, respectively. The 1-m-long tube sections were coupled together with coupling pieces made from the same material. Each coupling piece was equipped with one piston for compression of the sand, one valve for sampling of fluids during flooding, and one port for pressure measurement. Voids in the sand, generated by vibrations during sand filling, were taken up by moving the sand compression pistons. During packing, sand was repeatedly filled at each connection until the pistons could not be moved anymore. This assured that no voids were present at tube connectors. Flow ports and pressure ports were equipped with Hastelloy C-276 wire mesh in order to confine the sand. Each coupling piece changed the direction of flow by 180° such that the entire tube assembly only occupied a volume of 123×40×20 cm3 , and could be fitted into a thermostated cabinet. The layout of the apparatus is sketched in Fig. 1. The assembly was designed for a pressure limit of 620 bar at 90°C.
Core scale flooding experiments for relative permeability measurements were carried out in a single 1 m section.
All flooding experiments were carried out with the tubes oriented horizontally.
|File Size||195 KB||Number of Pages||6|