You are on page 1of 10

IBP1346_12 NUMERICAL SIMULATION OF AN ALTERNATIVE TO PREVENT HYDRATES FORMATION IN A BYPASS SECTION Lucilla C.

Almeida1, Joo Aguirre 2, Clarissa Bergman Fonte3, Fabricio Soares da Silva4, Carlos Alberto Capela Moraes5

Copyright 2012, Brazilian Petroleum, Gas and Biofuels Institute - IBP


This Technical Paper was prepared for presentation at the Rio Oi & Gas Expo and Conference 2012, held between September, 1720, 2012, in Rio de Janeiro. This Technical Paper was selected for presentation by the Technical Committee of the event according to the information contained in the final paper submitted by the author(s). The organizers are not supposed to translate or correct the submitted papers. The material as it is presented, does not necessarily represent Brazilian Petroleum, Gas and Biofuels Institute opinion, or that of its Members or Representatives. Authors consent to the publication of this Technical Paper in the Rio Oil & Gas Expo and Conference 2012 Proceedings.

Abstract
This work presents the use of Computational Fluid Dynamics to evaluate the feasibility of MEG (monoethylene glycol) injection as an alternative to prevent hydrate formation in a bypass section, present in an inlet module of a separation device of a subsea separation system. As the bypass section is open to the main pipeline, MEG will probably be dragged due to secondary flows generated by the main flow stream. The MEG removal rate is estimated, as well as the internal heat transfer between the currents and the heat loss to the external environment in order to estimate the temperature in the equipment. In a first step, the MEG removal was evaluated considering the heat transfer between the liquid phase (composed of water, oil and MEG) and the gas phase as well as the heat transfer by forced convection to the external environment. In a second step, the influence of a thermal insulation layer around the bypass line, reducing the heat loss to the external environment, was studied. Both simulations (with or without thermal insulation) showed the establishment of secondary flows in the open connection between the main line and bypass line, promoting the removal of MEG from the bypass section and enabling other components of the liquid phase and/or gas to enter in the bypass line. This MEG removal is faster when thermal isolation was considered, due to the fact that higher temperatures are established in the bypass, maintaining the liquid phase with lower densities and viscosities. With regard to temperature, the insulation was able to keep higher temperatures at the bypass line than those obtained without insulation, indicating that the combination of MEG injection and thermal insulation may be able to avoid the critical condition for hydrate formation.

1. Introduction
The inlet module of a subsea separation device, installed in a flowline downstream a production well, has a bypass section so that, in case of emergency, the production can be diverted. Therefore, this section must be unobstructed when its use is necessary. However, the low temperatures that occur on the seabed may induce the hydrates formation, which can damage or even block the flow through the bypass. It was proposed then to fill the bypass section with monoethylene-glycol (MEG), a fluid denser than water, in order to avoid hydrates in case of gas presence in this pipe section. However, as the bypass is open to the main pipeline where produced water, oil and gas flow, MEG will probably be dragged due to secondary flows generated by the main flow stream. Therefore, periodic replacement of MEG would be required. Another possibility to avoid hydrate formation would be the maintenance of the bypass section at temperatures higher than those that allow hydrate formation. Heating the pipe seems to be unfeasible, since the energy losses to the environment would be very high. However, as the produced current is under high temperatures (around 60 C), heat

______________________________ 1 Chemical Engineer ESSS 2 M.Sc., Mechanical Engineer ESSS 3 Chemical Engineer ESSS 4 Petroleum Chemist Petrobras 5 D.Sc., Mechanical Engineer Petrobras

Rio Oil & Gas Expo and Conference 2012 transfer combined with the recirculation at the entrance of the bypass may be enough to ensure the maintenance of reasonable temperatures in the bypass region. In this study the viability of this alternative (injection of MEG to ensure no hydrates formation in the section) was evaluated using CFD. The frequency with which the MEG is dragged was estimated, so that it is possible to calculate the necessary replacement frequency. In addition, the heat transfers between the internal currents and the heat loss to the external environment were studied in order to estimate the temperatures on the device.

2. Geometry
The geometry of the inlet module of the device was simplified in order to keep only the flow regions, as can be seen in Figure 1.

Figure 1. Studied geometry. In Figure 2, the selected simulation domain can be viewed in its original position. The red line is the separation device inlet line. The orange one is the by-pass line that ends in a block containing a valve.

Figure 2. Original Geometry (in orange bypass line, in red main line).

3. Mesh
2

Rio Oil & Gas Expo and Conference 2012 The studied geometry, shown in Figure 1, has been inserted in the ANSYS Workbench package and a grid of hexahedral elements was generated on it to solve flow using the finite volume method (in this work the ANSYS FLUENT software was used). The used mesh had approximately 700 000 elements and was refined in the near walls regions and in the contact region between the bypass line and the main pipe, as higher velocity gradients are expected at these regions. Figure 3 shows the generated mesh as well as the refinement near walls.

Figure 3. Mesh used in this study in detail the mesh refinement near walls.

4. Modeling
4.1. Flow pattern definition The module operates at a gauge pressure of 51 bar and 60 C. The produced current is composed by three fluids: water, oil and gas. Furthermore, as MEG (mixed with ethanol) is injected as an alternative for inhibiting hydrates, the presence of this additional fluid must be considered. As there was no prior knowledge of the flow pattern within the device, a study was conducted to estimate the flow regime on it. The map proposed by Taitel et al (1980) for gas-liquid two-phase flow in vertical pipes was used to identify the flow pattern. The determination of the flow regime was important to decide which model would be used to model the multiphase flow, as well as to choose the better boundary condition to prescript at inlet. Using the flow rates showed at Table 1, the flow regime obtained using the referred map was the disperse flow pattern, as can be seen in Figure 4. Table 1. Inlet flow rates. Inlet Flow Rates [m3/dia] Gas Oil Water 2185.0 835.5 2403.5

Rio Oil & Gas Expo and Conference 2012

Figure 4. Map used in the flow regime determination the red point shows the actual flow conditions. 4.2. Materials properties Water, oil and MEG were considered a single continuous liquid phase. The gas phase was treated as a dispersed phase. The properties of the individual materials (oil, water, MEG and gas) were prescribed by polynomials as function of temperature, as shown in Figure 5 for the MEG viscosity at 53 bar.
MEG viscosity (53 bar)
25,000 MEG + EtOH Poly. (MEG + EtOH) 20,000

Viscosity (cP)

15,000

10,000

5,000

y = 3E-07x4 - 0,0005x3 + 0,2525x2 - 59,004x + 5197 R2 = 0,9999

0,000 273

283

293

303

313

323

333

343

353

363

373

Temperature [K]

Figure 5. MEG viscosity versus temperature at 53 bar. The properties (density, viscosity, heat capacity, thermal conductivity) of the multicomponent liquid phase (water, oil, MEG) were locally calculated using mass fraction weighted mixture rules. 4.3. Multiphase modeling The Eulerian-Eulerian approach was chosen to model the multiphase flow inside the domain. Using this approach, mass, momentum, and energy transport equations were solved for each phase (liquid and gas phases). The liquid phase was regarded as the primary phase, while the gas phase was regarded as a dispersed phase, with a relatively large diameter (0.1 mm) in order to adequately represent the interfacial area concentration available for momentum and energy exchange between the phases. 4.4. Boundary conditions At the device inlet, the mixture velocity was specified, as well as the gas volume fraction. The gas is uniformly distributed and the MEG mass fraction is zero. Both phases enter the domain at a temperature of 60 C, with the same velocity. At the outlet face, zero relative pressure was prescribed.

Rio Oil & Gas Expo and Conference 2012 Non-slip condition was prescribed at walls for both phases (roughness equal to zero). The heat exchange by convection between the wall and the external fluid was considered (the heat transfer coefficient was prescribed at the walls). 4.5. Heat transfer coefficient calculation The line is exposed to the ocean currents, which in turn are in much lower temperatures than the produced fluids. Thus, heat is transferred by forced convection between the walls and the sea, considering its current. As the fluid properties are functions of temperature, it is important to take into account this heat loss during simulation. Experimental evaluations (Handbook of Single-Phase Convective Heat Transfer, 1987) show that the Nusselt number calculated for a flow around a cylinder varies with the angle to the stagnation point, since heat transfer is greatly influenced by how the boundary layer develops on the surface. However, at this work, global conditions are needed. A known correlation, given by Zukauskas (1987), was used to calculate the heat transfer coefficient by forced convection for the flow around a cylinder, according to Equation 1:

Pr Nu C Re Pr Pr s
m n

1/ 4

(1)

The Reynolds and Prandtl numbers are evaluated at T , except PrS, which is measured at surface temperature. The equation below is used to correlate the Nusselt number and the heat transfer coefficient, making it possible to prescribe the heat transfer coefficient as a function of local wall temperature:

Nu

hD kf

(2)

A C routine was written, returning the heat transfer coefficient for each element of the wall, for each time step, as a function of the wall temperature. The calculation is passed to the setup using a UDF (User-Defined Function). 4.6. Conduction through the walls To account for conduction through the walls without the need to mesh them, which could lead to a mesh with high aspect ratio and a significant increase in cell number, a heat conduction equation was added to account for the thermal resistance afforded by the wall. Thus, the heat conduction through the wall thickness will be regarded as well as the conduction within the wall. The thermal resistance of the walls is

x k , where k

the conductivity of the material composing the wall and

x is the wall thickness. Using this approach, the thermal boundary condition specified for the wall (the heat transfer coefficient) will be prescribed at the outside surface of the wall, ie, the wall opposite the fluid face, as shown in Figure 6.

Rio Oil & Gas Expo and Conference 2012

Figure 6. Thermal boundary condition at walls 4.7. Turbulence modeling The RANS modeling was used, in which the Navier-Stokes equations are time averaged and the arising terms are modeled using the turbulent viscosity concept, according to Boussinesq hypothesis. The standard - model, which is based on the modeling of the turbulent kinetic energy and turbulent dissipation rate, was used, with the enhanced version of the wall function treatment. 4.8. Initial conditions The domain was initialized with a prescribed level of MEG, as expected immediately after the injection. Thus, part of the domain had liquid volumetric fraction equal to 1, and MEG mass fraction equal to 1. Outside it, gas and liquid were uniformly mixed, and the MEG mass fraction in the liquid mixture was equal to 0, as can be seen in Figure 7. The fluids were at 60 C while the average temperature of the sea current is 4 C.

Figure 7. MEG distribution at the begin of the simulation

5. Results
The methodology to calculate the heat transfer coefficient for each element of the wall as a function of temperature produced consistent results. As can be seen in Figure 8, there is heat transfer between the external fluid, that has a fixed temperature of 4 C (sea water), and the pipe wall. Thus, the temperature established at each element of the wall is function of the heat exchanged between the sea and the fluids studied. 6

Rio Oil & Gas Expo and Conference 2012

Figure 8. Wall temperature. As the region closest to the bypass valve is almost stagnant, the heat transfer between the wall and the produced fluid plus MEG is practically entirely due conduction and the wall temperatures are lower than those near the region where the open connection between the main line and the bypass line and recirculations take place.

Figure 9. Isosurfaces of constant MEG mass fraction. Isosurfaces of constant MEG mass fraction were tracked over time, as shown in Figure 9. MEG is removed of the bypass section due to recirculations that are established at the open connection between the main line and the bypass line. The isosurface of MEG mass fraction equal to 0.95 gradually retreats towards the valve, indicating that other components of the liquid mixture or gas enter the bypass section, which will no longer contain only MEG as in the beginning of the simulation. The reduction of the MEG filled region is faster in the case with thermal insulation than in the case without insulation, since higher temperatures are established in the bypass, maintaining the liquid phase with lower densities and viscosities, facilitating MEG removal due drag. As the combination of gas presence and temperatures below 15C may lead to hydrates formation, according to hydrate formation simulation at considered pressure, the bypass section was separated into three consecutives regions, as shown in Figure 10 , in order to evaluate how much gas enter at these regions and how the temperature decays. 7

Rio Oil & Gas Expo and Conference 2012

Figure 10. Bypass divisions. The average temperatures at each region for the cases with and without insulation are shown in Figure 11. The heat exchange with the external environment causes the decrease of the average temperature inside the bypass. This decrease is less significant in the case with insulation, showing that the insulation is able to limit the heat losses to the external environment, holding the bypass section at higher temperatures than in the case without insulation.

Figure 11. Average temperatures versus time with and without insulation Similarly, the average gas volume fraction was evaluated in each region. As shown in Figure 12, gas reaches all regions, including the region near to the valve. The gas invasion occurs faster in the case with insulation than in the case that no insulation is provided. The fluids viscosities and densities are smaller in the case with thermal insulation, which has higher temperatures compared with the case without isolation, making easier the MEG removal and therefore gas penetration in the bypass section.

Rio Oil & Gas Expo and Conference 2012

Figure 12. Gas volume fraction versus time with and without insulation Temperature profiles for the gas phase were measured over several sectional planes, as shown in Figure 13. For both cases, gas tends to accumulate in the region closest to the top of the pipe (due to the density difference between liquid and gas phases). For the case without insulation, there are regions where gas exists with temperature lower than that allows hydrate formation. However, for the case with insulation, the temperature is maintained above 15C throughout the bypass. Thus, the combination of continuous MEG removal due to secondary flows in the open connection between the main pipe and the bypass pipe, and thermal insulation, restricting heat exchange with the external environment, may be able to avoid critical conditions to hydrate formation.

Figure 13. Gas temperature profiles no insulation case

6. Conclusions
The study consisted of the analysis of the feasibility of MEG injection in order to avoid hydrates formation at a bypass section. The heat transfer between the produced fluids and the injected MEG, and the heat loss to the environment were evaluated in order to estimate the temperature inside the bypass section, with and without thermal insulation. For both cases (with and without thermal insulation), the secondary flows that are established in the open connection region between the main line and bypass line cause MEG removal of the bypass section due to drag. 9

Rio Oil & Gas Expo and Conference 2012 Isosurfaces of MEG mass fraction equal to 0.05, 0.5 and 0.95 were followed throughout the simulation. These gradually recede toward the valve, indicating that other components of the liquid mixture and/or gas enter in the bypass section. Thus, the bypass section is no longer filled only by MEG as in the beginning. The MEG removal is faster in the case with insulation than in the case in which no insulation is provided. This is due to the fact that higher temperatures are established in the bypass, maintaining the liquid phase with lower densities and viscosities, facilitating entrainment of MEG. In both cases, gas penetrates through all the bypass section, achieving the region closer to the valve, further away from the open connection between the main line and the bypass line. With regard to temperature, the insulation was able to keep the higher average temperatures inside the bypass than in the case in which pure steel was considered as the wall material. During the analyzed time (about 13 min), average temperatures throughout the domain were above the critical temperature for hydrates formation in the case with thermal insulation. Thus, the combination of continuous removal of MEG due to secondary flows in the open connection of the main pipe and the bypass pipe and the thermal insulation, restricting heat exchange with the external environment, may be able to avoid critical conditions that allow the onset of hydrate.

7. Acknowledgements
The authors express their sincere acknowledgements to PETROBRAS for giving its permission to publish this article, and giving the opportunity and resources that made this study possible.

8. References
An Introduction to Computational Fluid Dynamics: The Finite Volume Method, Versteeg, H. K, Malalasekera, W., Pearson Education Limited, 2007. ANSYS-FLUENT Theory Guide (2010), Release 14.0. ANSYS Inc. Boundary-Layer Theory, Schlichting, H., Gersten, K., Springer, 2000. CROWE, C. T. Multiphase Flow Handbook. Taylor & Francis Group, 2006. FONTE, C. B. Estudo de mapas de regimes de escoamento gs-lquido em dutos horizontais, Projeto de concluso de curso, Escola de Qumica, UFRJ, 2010 GOVIER, G., AZIZ, K. The flow of complex mixtures in pipes. Nova Iorque:Van-Nostrand Reinhold, 1972 GOVIER, G. W., RADFORD, B.A., DUNN, J. S. C, The upwards vertical flow of air-water mixtures , Can. J. Chem. Eng., 35, 5870, 1957 Handbook of Single-Phase Convective Heat Transfer, 1987. JONES, W.P., LAUNDER, B. E., The Prediction of Laminarization with a Two-Equation Model of Turbulence, International Journal of Heat and Mass Transfer, vol. 15, 1972, 301-314, 1972. LAUNDER, B.E., SPALDING, D.B., The numerical computation of turbulent flows, Computer Methods in Applied Mechanics and Engineering 3 (2), 269-289, 1974. TAITEL, Y., BORNEA D. and DUCKER A. E., Modelling Flow Pattern Transitions for Steady Upward Gas-Liquid Flow in vertical Tubes, AIChE Journal, 26, (3), 345-354, 1980.

10

You might also like