(149l) Harnessing Feedback Control Strategy for Improved Core Annular Flow Stability in Heavy Oil Transportation

The Core Annular Flow (CAF) pattern, also known as annular flow, is a phenomenon that occurs in multiphase flow systems where one fluid is encircled by another external fluid within a pipeline. This flow pattern is crucial in various engineering fields, particularly in the oil and gas industry¹.

One of the primary applications of CAF lies in the transportation of viscous fluids. In such cases, it is essential to utilize an external fluid with low viscosity to minimize the friction between the internal fluid and the pipeline wall. Commonly, water serves as the less viscous fluid, resulting in a liquid-liquid system². A visual representation of this system is depicted in Figure 1, where high-viscosity oil is at the center, and water surrounds it, functioning as a lubricant.

Figure 1 – Example of Core Annular Flow.

By utilizing water as an external fluid, the application of CAF enables more efficient oil transportation with reduced pressure loss. This is particularly beneficial in regions with cold climates, where oil viscosity tends to increase³.

Despite being used in large-scale transport systems, CAF still faces various challenges. These include destabilization of annular flow due to geometric variations or changes in the pipeline's flow direction, fouling caused by viscous fluids adhering to the pipeline wall, emulsion formation resulting from solid impurities or water present in the fluid that may impact the annular flow pattern's stability, and maintaining CAF stability over extended distances due to the buoyancy effect.².

Control solutions are often the first consideration when addressing stability issues in other engineering problems. However, the authors are unaware of any existing literature specifically tackling control issues in CAF systems.

On the other hand, there are notable applications of PID control in addressing oil and gas transport challenges, as reported in Jahanshahi et al. (2012)⁴. In this study, the authors aimed to develop a simple and robust control structure for stabilizing gas-lifted oil wells, preventing casing-heading instability. They performed a controllability analysis using various candidate control variables and manipulated variables, ultimately creating an effective control structure based on top-side pressure measurements. In another study by Jahanshahi et al. (2013)⁵, PI and PID controllers were employed to prevent unstable flow in offshore oil fields. The authors proposed robust and applicable tuning rules through model identification and IMC design, validating their results using test rigs and simulations with the OLGA simulator.

Therefore, this work intends to utilize CFD (computational fluid dynamics) techniques to simulate a CAF system and subsequently develop a PI strategy for controlling this process. By increasing the spatial occupation of oil in the pipe's cross-section, the controller will seek to maintain the process in a highly productive region while preventing fouling in the presence of disturbances.

METHODOLOGY

Building upon the CAF induction head concept, which features water inlets at the radial extremity of the pipe, CFD simulations were carried out using Ansys Fluent software. An isothermal system was assumed, in which the physicochemical properties of the fluids remain constant.

For the simulation, densities of 998.2 kg/m³ and 854 kg/m³ were considered for water and oil, respectively. The viscosities were set at 0.001003 Pa·s for water and 0.62 Pa·s for oil. The interfacial tension between water and oil was 0.032 N/m⁶. The multiphase system employed the VOF (Volume of Fluid) model, while the RANS 𝐾−𝜖 SST model was used for viscosity, with refined treatment on the walls.

A two-dimensional, 1-meter pipe was simulated, featuring a 50 mm inlet. The central oil inlet measured 40 mm, accompanied by two peripheral 5 mm water inlets. Figure 2 illustrates the generated geometry representing the induction head.

Figure 2 - Geometry for the induction of the Core Annular Flow.

Upon defining the system, open-loop simulations were conducted, varying the boundary conditions (water and oil velocities) to assess their impact on the spatial occupation within the pipeline. Initially, the pipe was filled with water. Subsequently, closed-loop tests were initiated using MATLAB software to implement a proportional controller in the system. This aimed to control the spatial occupation of oil within the pipeline and evaluate the controller's ability to withstand disturbances in the system.

RESULTS

Through open-loop simulations of the system, it became apparent that the oil's spatial occupation could vary significantly. Moreover, it was observed that certain operating conditions might be unstable and could potentially lead to fouling within the pipeline in the presence of disturbances. Some of these phenomena can be seen in Figure 3, emphasizing the ratio 0.35 and ratio 0.1, which begin to display foul.

Figure 3 - Simulations of different operating conditions.

The figure above shows that a decrease in the water injection rate leads to an increase in the volumetric fraction occupied by the oil. This effect is noticeable when comparing ratio 2 with ratio of 1.5, where a slight increase in oil volume can be observed. As the water-to-oil rate is reduced, the buoyancy effects on the oil become more significant due to the oil's lower density than water. Consequently, the volume occupied by the oil tends to increase in the cross-section up to a critical point where a reduction in water velocity results in pipe fouling.

A proportional controller with a gain of 0.25 was proposed to address the control challenges in the CAF system. The controller's objective is to track the oil fraction setpoint by manipulating the water velocity while considering the oil velocity as an unmeasured disturbance.

The proposed control scheme is expected to improve the CAF system's stability, particularly in the presence of disturbances, by controlling the oil fraction. The controller can prevent fouling deposits on the pipeline wall and maintain the system's productivity.

To evaluate the effectiveness of the proposed control scheme, closed-loop simulations were conducted. The system started with a high fraction of oil, and the value of 0.6 was established as a setpoint, defined from the open loop stability limit. In addition, for every 15 actions, there is an unmeasured disturbance of 0.025 m/s in the oil velocity, making a step from 0.3 m/s to 0.375 m/s in oil speed, as seen in figure 4.

Figure 4 - Result of controlling the oil fraction in the pipeline cross-section.

Figure 4 shows that the controller successfully maintained the setpoint for disturbance rejection. Comparing it with Figure 3 reveals that the second disturbance would have caused the system to enter a stratification regime, leading to fouling the pipe’s wall. To prevent this, the controller increased the water velocity, ensuring the CAF remained in the annular flow regime. It is remarkable that a single feedback loop was enough to keep the system under the annular regime, making it possible to improve the system in future works with more advanced control strategies.

CONCLUSIONS

In conclusion, the Core Annular Flow (CAF) pattern is a significant phenomenon in multiphase flow systems. Despite its wide-ranging applications, challenges in maintaining the stability of the annular flow pattern remain. Using computational fluid dynamics (CFD) simulations and the implementation of PID control techniques, this study demonstrates the potential for enhanced system performance in CAF systems for transporting heavy oil in liquid-liquid systems. The results indicate that the proportional controller successfully controlled the spatial occupation of the oil in the pipeline and demonstrated resistance against disturbances in the system. This research highlights the importance of understanding the dynamics of CAF systems and presents an opportunity to apply control techniques for improved efficiency and stability.

REFERENCES

¹ Rosa, E. S. (2012). Escoamento multifásico isotérmico: modelos de multifluidos e de mistura.

² Joseph, D. D., Bai, R., & Chen, K. P. (1997). CORE-ANNULAR FLOWS. 65–90.

³ Tripathi, S., Tabor, R. F., Singh, R., & Bhattacharya, A. (2017). Characterization of interfacial waves and pressure drop in horizontal oil-water core-annular flows. Physics of Fluids, 29(8). https://doi.org/10.1063/1.4998428

⁴ Jahanshahi, E., Skogestad, S., & Hansen, H. (2012). Control structure design for stabilizing unstable gas-lift oil wells. IFAC Proceedings Volumes (IFAC-PapersOnline), 8 (PART 1), 93–100. https://doi.org/10.3182/20120710-4-SG-2026.00110

⁵ Jahanshahi, E., & Skogestad, S. (2013). Closed-loop model identification and PID/PI tuning for robust anti-slug control. IFAC Proceedings Volumes (IFAC-PapersOnline), 10(PART 1), 233–240. https://doi.org/10.3182/20131218-3-IN-2045.00009

⁶ Li, X. G., Zhang, L. H., Zhang, R. Y., Sun, Y. L., Jiang, B., Luo, M. F., & Li, X. G. (2015). CFD modeling of phase change and coke formation in petroleum refining heaters. Fuel Processing Technology, 134, 18–25. https://doi.org/10.1016/j.fuproc.2015.03.005

⁷ Li, L., Popa, F., & Houchens, B. C. (2015). Mechanistic prediction of oil-water, two-phase flow in horizontal or near-horizontal pipes for a wide range of oil viscosities. Proceedings - SPE Annual Technical Conference and Exhibition, 2015-January, 64–78. https://doi.org/10.2118/174726-ms