|  homepage  |  news  |  contact form  |  site map  | 

Flow Modelling of a FCC Slide Valve
for Oil Refinery Systems - 3

Flow Field Simulation

The valve flow field has been simulated using a package based on finite volume methods. First of all, a two-dimensional case study was performed to verify the approach to follow in the three-dimensional one. Both cases were considered as a single-phase flow problem. Later a two-phase study was performed for a two-dimensional geometry.

Three-Dimensional Study

Thanks to the valve symmetry plane, it was possible to model just one half of the valve. In order to reduce the number of cells, a short geometry at the valve outlet was designed by using a pressure boundary condition instead of a velocity one. The number of cells was 36 × 22 × 58. The required pressure drop through the valve was used as input in the port calculation program. In order to have a 13 kPa pressure drop the spreadsheet program calculated an opening of 64% for the valve disc.

These values have been used for the case inputs:

• Type of fluid: Isothermal, single phase;
• Mass flow rate: 502 kg/s;
• Dynamic viscosity: 2.5 Pa s;
• Density: 525 kg/m³;
• Valve opening: 64% of total area;
• Boundary conditions: pressure drop = 13 kPa.

It is worth mentioning that the flow field has been calculated by using the pressure boundary condition and valve opening according to the spreadsheet program. It was possible to see an erosion problem on the restriction walls before the outlet. The plotting of the static pressure confirmed that the fluid is impacting on the lower side: the kinetic energy of the fluid is transformed into potential energy. The jet-like flow coming out of the valve opening [image] and the deflection of the jet toward the side wall, which is dangerous for erosion, can be avoided with improved design of slide valves. An effort should be made to move the working valve opening i.e. 64% of total port opening, toward the outlet centerline, or better, a bit further. With excluding the steam injection, this study proved that the fluid in the bonnet has very little kinetic energy and no erosion has to be expected in that zone.

Two-Phase Study

A two-phase study was performed for a two-dimensional geometry with 79 × 39 cells. The solver used for a two phase flow needs to work with a time dependent approach, thus a very large number of iterations is necessary before reaching a steady state solution. This is why a three-dimensional approach was ruled out: calculation time would have been way too long.

With this approach it is not possible to use pressure boundary conditions, so an average velocity was calculated for the input in order to respect the mass flow rate. An outlet boundary condition has been used for the valve outlet; because of this, the geometry after the wall restriction at the valve exit had to be increased in length in order to stabilize the fluid flow before the outlet section.

The case data are the following:

• Type of fluid: Isothermal, two phase, granular flow;
• Volume fraction: gas = 0.6026, catalyst = 0.3974;
• Mass flow rate: 502 Kg/s;
• Gas viscosity: 4.6 e-5 Pa s;
• Gas density: 0.78 kg/m³;
• Particles density: 1320 kg/m³;
• Valve opening: 64% of total area;
• Outlet boundary condition: Outlet;
• Boundary conditions for the main inlet: Vx = 2.4 m/s; gas volume fraction = 0.6026;
• Boundary conditions for the purge inlet: Vy = 0 m/s, gas volume fraction = 1; Vy = 15 m/s, gas volume fraction = 1; Vy = 200 m/s, gas volume fraction = 1;
• Body force: gx = 7.514 m/s (40° valve inclined).

First of all, a steady state condition was reached without steam purge injection. A plotting of the particles' velocity is shown in figure 2. It is possible to notice a negligible difference between this and the three-dimensional flow fields. This fact proves the possibility to use a single phase model to solve these types of problems.

It is noticeable that different conditions have been used for the two cases.

The purge gas injection has been activated to verify the erosion of the disc stem, due to the impact of catalyst particles energized by the the purging steam, starting from the steady state solution.

A first case was run with a gas velocity of 15 m/s and a second one with 200 m/s.

In order to estimate the erosion on the disc stem, the catalyst vectors' magnitude, corresponding to the stem borders, has been multiplied by the function for erosion for ductile materials (ASME, Centennial Research Project, 1980), shown in figure 3. Results have been normalized to one.

The graph below shows the estimated erosion along the stem borders.

Conclusions

The results obtained through this study stress the importance of some key factors:

• the internal structure of the valve body should be designed to avoid a direct impact of catalyst particles on the walls;
• dimension of the calibrated orifices on the stuffing box purging line must be calculated with extreme precision. The wrong sizing would cause serious damage to the internal parts - disc, shaft and guides.

back to CFD part 2

Copyright © REMOSA spa 1998-2001 Last updated on May 18, 1998