Scaled Physical Modeling of Steam-Injection Experiments
- K.D. Kimber (Alberta Research Council) | S.M Farouq Ali (U. of Alberta)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- November 1991
- Document Type
- Journal Paper
- 467 - 469
- 1991. Society of Petroleum Engineers
- 5.8.5 Oil Sand, Oil Shale, Bitumen, 5.2 Reservoir Fluid Dynamics, 2.4.3 Sand/Solids Control, 5.4.6 Thermal Methods, 5.5 Reservoir Simulation, 4.6 Natural Gas
- 1 in the last 30 days
- 347 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
Summary. Scaled model experiments are valuable aids to steamflood design and numerical simulation. Current scaling techniques, however, have limitations. Because the techniques require porous media and pressure conditions different from those in the field studied, pressure-dependent properties and porous media characteristics are not scale accurately. This properties and porous media characteristics are not scale accurately. This paper discusses scaling techniques that allow field porous media and paper discusses scaling techniques that allow field porous media and pressure conditions to be used. The ability of scaling approaches to pressure conditions to be used. The ability of scaling approaches to predict such parameters as oil production rate, temperature distribution, predict such parameters as oil production rate, temperature distribution, and pressure conditions is considered. The new approaches provide scaling options for steam processes that use additives (PVT properties must be scaled), that use pressure cycles (fluid compressibilities must be scaled), and that require use of the same porous medium as found in the field.
Scaled models of steam-injection processes continue to be of great value in understanding complex mechanisms. To study processes where additives or pressure cycles are used with steam, high-pressure models that use reservoir fluids and reservoir pressure and temperature conditions are required. Endpoint saturations and relative permeabilities are scaled better if the porous medium used in permeabilities are scaled better if the porous medium used in the models is the same as that in the prototype. New scaling approaches, designed to satisfy these conditions, were developed previously by Kimber et al. previously by Kimber et al. Experiments Conducted
The objectives of the experimental investigation were (1) to assess the relative ability of each scaling approach to predict important parameters in steam-in action processes, (2) to evaluate the parameters in steam-in action processes, (2) to evaluate the importance of scaling pressure conditions and the significance of what is relinquished to do so, and (3) to verify that the novel concepts developed for the new scaling approaches were valid. Three scaling approaches were examined. Approach 1, originally developed by Pujol and Boberg, satisfies the scaling requirements for geometry and gravitational effects but requires porous media and pressure conditions different from those found in the field. Approach 2 relaxes scaling of the gravitational effects to satisfy scaling of geometry and pressure conditions. It also allows prototype porous media to be used in the model. Approach 3 satisfies scaling porous media to be used in the model. Approach 3 satisfies scaling of both the pressure conditions and the gravitational effects while using prototype porous media, but relaxes some aspects of geometric similarity. These approaches were tested with steam/water/sand systems, and the results were reported in 1989. The present study extends their verification to the steamflooding of oil- and water-saturated porous media by use of a high-pressure scaled model. Details of the experimental work are given in Ref 4. Dimensionless experimental results from a large model (prototype) were compared with those from small models, which were scaled down from the large model with each of the three approaches. To ensure a realistic combination of forces, geometry, boundary conditions, and material characteristics, the large model was designed by scaling down one-eighth of a field five-spot pattern. The results for the steamfloods in the large- and small-model experiments were then compared to determine the ability of each scaling approach to predict the results of the large model. The large model was only twice (linear scale) the size of the smaller models, as necessitated by experimental considerations. The experiments used a 30-mPas [30-cp] oil and water-wet unconsolidated sand.
Predictions of the large-model results based on Approaches 1 Predictions of the large-model results based on Approaches 1 through 3 were compared for a number of different dimensionless parameters, including production rates, temperature distribution, parameters, including production rates, temperature distribution, and pressure profiles.
Production. Accurate prediction of the oil production rate is an Production. Accurate prediction of the oil production rate is an important feature of any scaling approach. Production rate, rather than cumulative production, was compared because it is more sensitive to steam breakthrough and maximum steam-zone development (Fig. 1). Approaches 1 and 2 make very good predictions of the large-model results, while Approach 3 is less accurate. The oil production rates are high as the steam zone develops and drop off production rates are high as the steam zone develops and drop off rapidly as the oil/water ratio declines.
Temperature Distribution. Scaling of the temperature distribution is critical in any effort to scale a thermal recovery process. The ability of the small model to predict the results of the large model depends on the scaling of the thermal properties of the fluids and rock as well as the fluid distributions. Vertical profiles give an indication of the effects of gravity override, heat losses to the surrounding formations, and the ability of each scaling approach to scale these mechanisms. Fig. 2 compares the temperature profiles for vertical cross sections through the center of the profiles for vertical cross sections through the center of the overburden, reservoir, and underburden. Only the comparison between the large model and Approach 2 is shown, but predictions made by the other approaches were of similar quality. Comparisons made at other times during the experiment also agreed closely. Even though Approach 2 relaxes scaling of gravitational effects, the override evident in the large model is well matched with this approach. The override is more severe in these experiments than for the steam/water/sand experiments reported previously. The temperature profiles in the overburden and underburden indicate proper scaling of heat transfer to these regions. Approach 3 gave proper scaling of heat transfer to these regions. Approach 3 gave the least accurate predictions of energy losses to the surrounding formations, as it did for the previous steam/water/sand experiments.
Pressure Data. One of the motivations for tills study was to develop Pressure Data. One of the motivations for tills study was to develop scaling criteria with improved scaling of the pressure conditions and therefore fluid properties that depend on pressure. Fig. 3 shows a comparison of the dimensionless injection pressures for the large model and those predicted by each of the three scaling approaches. Very good predictions were made by Approach 2, fair predictions by Approach 3, and rather poor predictions by Approach 1. This is expected because the major weakness of Approach 1 is its inability to scale the pressure conditions and any dependent properties. When a process is scaled in which pressure conditions properties. When a process is scaled in which pressure conditions and corresponding fluid properties change significantly, Approaches 2 and 3 yield better results.
1. Reasonable predictions of process parameters were achieved with Approaches 1 through 3 in the steam/oil/water system studied. Approaches 2 and 3 offer new scaling options for design of steam-injection experiments. 2. Approaches 2 and 3 yielded more accurate scaling of the pressure conditions and therefore the fluid properties that depend on pressure conditions and therefore the fluid properties that depend on pressure. This is critical to the extension of scaling beyond basic pressure. This is critical to the extension of scaling beyond basic steamflooding.
|File Size||190 KB||Number of Pages||3|