Theoretical and Experimental Investigation of Isothermal Compositional Grading
- J. Ratulowski (Schlumberger Canada Ltd.) | A.N. Fuex (Shell Global Solutions) | J.T. Westrich (Shell Intl. E&P B.V.) | J.J. Sieler (Shell Intl. E&P Inc.)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- June 2003
- Document Type
- Journal Paper
- 168 - 175
- 2003. Society of Petroleum Engineers
- 5.2.2 Fluid Modeling, Equations of State, 1.7.5 Well Control, 2.4.3 Sand/Solids Control, 5.2 Reservoir Fluid Dynamics, 1.10.1 Drill string components and drilling tools (tubulars, jars, subs, stabilisers, reamers, etc), 4.3 Flow Assurance, 5.2.1 Phase Behavior and PVT Measurements, 4.3.3 Aspaltenes, 4.1.5 Processing Equipment, 4.3.4 Scale, 5.5 Reservoir Simulation
- 2 in the last 30 days
- 474 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
A centrifuge system capable of producing a potential difference across a live oil column equivalent to 1,000 ft of gravitational head was designed and tested. Initial tests on a simple ternary system yielded results similar to equation-of-state (EOS) models. A blackoil sample from a Gulf of Mexico field, Bullwinkle J2-RB sand, exhibiting compositional gradients was segregated in the centrifuge. Experimental results from the centrifuge were similar to field values, indicating that the large variation in composition observed for this field may be attributed to gravitational segregation alone.
From the results of these two sets of experiments and the subsequent analysis of the graded fractions, we can conclude that significant compositional variation of reservoir fluids not near their vapor/liquid critical points can be caused by gravity alone. The grading phenomenon is sensitive to the saturate/aromatic balance of the oil. Existing EOS models did not correctly predict the compositional variations in fluids like the Bullwinkle J2-RB because pseudocomponents generated from volatility distributions have a fixed saturate/aromatic character. It is subtle changes in the saturate/aromatic balance that drive the grading phenomena for these fluids.
Compositional variation with depth has been observed in many reservoirs. These gradients result from a variety of causes and typically indicate nonequilibrium states. Gradients in composition can be observed in systems in equilibrium when chemical potential gradients are balanced by the gravitational potential gradient.1-8 Temperature gradients can also contribute to concentration variation. However, in this situation a steady state (not an equilibrium state) exists in the reservoir.1,9,10 Equilibrium composition gradients have been observed and modeled by various authors in nearcritical reservoirs. Whitson1,2 provides a comprehensive summary of these efforts.
We have encountered a reservoir in the Gulf of Mexico, Bullwinkle J2-RB, with significant variation in gas/oil ratio (GOR) and saturation pressure with depth. The Bullwinkle fluid is not near its vapor/liquid critical point at any location in the column. However, the saturation pressure gradient is 3 psia/ft, and the GOR changes by nearly 700 scf/bbl in 800 ft of oil column. The gradient in saturation pressure is even larger than equilibrium gradients typically predicted for near-critical fluids. Existing EOS models could not be tuned to fit this behavior. Analysis of the gas shows no significant variation in methane isotope ratios across the column. This indicates that at least the light ends are well mixed, and it suggests an equilibrium condition in the reservoir. Other geochemical data, including fingerprinting, support this conclusion.11
To develop tools to predict composition variation away from well control, we need to understand the mechanism causing the gradients in fluid properties in the Bullwinkle J2-RB. An experimental apparatus to test the gravitational grading mechanism at Bullwinkle was designed. The results of these experiments and additional modeling work have provided us with a mechanism that can account for the observed composition variation in the Bullwinkle J2-RB.
Variation in gravitational potential across a 1,000-ft hydrocarbon column is simulated by centrifuging a vessel of live fluid under pressure in the laboratory. The body force generated by the centrifugal field is analogous to that generated by a gravitational field. The rotation speed of the centrifuge is chosen such that the change in potential across the vessel is approximately the same as the change in gravitational potential across a 1,000-ft hydrocarbon column. For the geometry of our experiment, the rotation speed required is 5,000 rpm. The design requirements for both the pressure vessel and the centrifuge are challenging. To grade and analyze fluids at reservoir conditions, the pressure vessel must have 10,000 psia working pressure; 100°C working temperature; 2,500 g maximum total weight of vessel and contents (centrifuge specs); volume large enough for multiple, precise measurements of fluid properties and GOR; and two valves and associated fittings that will not leak under the test conditions. The pressure vessel that is shown in the spinning configuration in Fig. 1 meets all these requirements. Because of safety concerns, the centrifuge is operated in a barricade by remote control. This experiment uses the maximum capability of the centrifuge in terms of speed and mass being spun. Also, the fluid being centrifuged is highly volatile, and leakage would create conditions unacceptable for a normal laboratory environment. In all the experiments to date, we have noticed no significant change in pressure caused by leakage in the vessel before and after spinning.
The sample is charged isobarically and isothermally into the sample vessel through the central tube shown in Fig. 1. Mercury is removed from the bottom. The centrifuge is a slightly modified commercially available model, Beckman J6-HC. As shown in Fig. 1, the sample is contained in a small pressure vessel. The sample vessel replaces a bucket that is standard to the centrifuge. The bucket directly opposite the one containing the sample cylinder is replaced with an adjustable counterweight. The counterweight and the filled vessel are carefully balanced before the start of the experiment. The vessel swivels such that during spinning, the centrifugal force holds the bucket against a stop at the angle shown in Fig. 1.
|File Size||1 MB||Number of Pages||8|