In case you missed it, there is an amazing video of a CFD study of a FCC regenerator on the web. Software by Computational Particle Fluid Dynamics (CPFD) was used. Here's a link to the video on their website. In addition, a paper by Clark et al. describes the model and the results. I will be referring to both the video and the paper in this blog.
Having built CFD models 20+ years ago, I am in awe of this model. The capabilities have really increased since that time. Please view the video and paper before reading on.
Here's what I see
The afterburn problem
As Clark states, one of their goals for the model was to
"predict the severity of afterburn in the freeboard, and understand its causes". Based on the results from my simple model, it appears that this reactor exhibits both of the system causes of high afterburn...(1)low bed height and (2) high bypass of air.
What might be done to improve performance?
Reduce catalyst size
I think the main problem is the catalyst fluidization behavior. That might be improved with smaller particles. Improving the fluidization should reduce the amount of bypass and spouting. Smaller particles would also allow less air to be used to convey the spent catalyst from the stripper. This might allow better distribution of air through the air rings.
Increase bed height
The current reactor is nearly all freeboard with plug flow. The plug flow nature and the low heat capacity (low catalyst fraction) in the freeboard causes the high flue temperature. More burning needs to occur in the dense bed.
In addition to the benefit of higher catalyst residence time in the dense bed, increasing the bed height may also reduce the amount of bypassing. The added height would increase the pressure drop through the bed and possibly lead to more chaotic bubbling flow as opposed to jet flow. The added height of dense emulsion might also better absorb some of the high momentum of gas and catalyst entering from the distributor arms.
The bed height can be increased by raising the standpipe.
The air distribution problem
The need to provide sufficient air to convey the spent catalyst from below the reactor removes flexibility for distributing the air. Also, even though the distributor arms have 192 nozzles, the CFD simulation indicates that the air from an arm tends to converge due to the partially vertical nature of the arms. Changing the air distribution will be difficult. Catalyst particle size and bed height are relatively easy changes that may help reduce the gas bypass problem.
Do I need to change my freeboard model?
Currently my model assumes both the gas and the catalyst are in plug flow. If I were to change the catalyst to CSTR behavior, what temperature would I use? With my current model there is significant axial mixing because I am using only 5 tanks in series for the freeboard region. I think this is adequate.
CFD vs. the mixed model method
My model 2.1 showed potential causes of afterburn, even without modeling this reactor. If a tracer study had been performed on the reactor, it would have quantified the amount of bypass flow and shown the high catalyst fraction in the dense bed. The low bed height would have been known a priori from the geometry. The high catalyst fraction in the bed would be an indication of poor fluidization.
My simple model would not show that the poor fluidization may be partially due to the catalyst size or type. I think the CFD model more clearly shows this catalyst problem. The CFD model also shows the lack of movement in the dense bed and the hot spots which a gas tracer study would not reveal. There have been tracer studies for the catalyst phase using tagged particles, but the void fraction may be enough indication of the problem.
The CFD model appears to validate the general assumptions and findings of the mixed model method. Thus, the mixed model should be an adequate representation for optimization studies.
Clark, Samuel M, CPFD Software, and Dale M Snider. “Multiphase Simulation of a Commercial Fluidized Catalytic Cracking Regenerator FCC Regenerator As-Built System Geometry Multiple-Simulation Strategy for the Regenerator System Spent Catalyst Riser Simulation,” n.d., 1–9.
available on CPFD website