Scale-up of bubble columns: Controlling the effect of scale using internals

CRELChemical Reaction

Engineering Laboratory

The hydrodynamics of bubble columns are strongly affected by the scale of operation.

While most research has been done in columns of about 25 cm, the industrial units range from 6-10 m in diameter (Krishna et. al., 2001).

Scaling rules via geometrical and dynamic similarities are hard to achieve in such complex systems.

Novel approaches are needed to solve the scale-up quest.

BackgroundIn 1958, Kölbel and Ackermann patented a slurry

reactor design for carrying out the Fischer-Tropsch process.

Their design was meant to decrease the disadvantageous recirculating effect well known to occur in large commercial scale columns.

Subdividing the reactor space using similar vertical shafts open at both the top and the bottom was proposed to overcome the consequences of circulation.

CHEMICAL REACTION ENGINEERING LABORATORY

The ultimate result of the work by Kölbel and Ackermann was the elimination of the large vertical recirculation loops and the formation of a stable liquid-gas suspension system in each shaft with uniform sizes of gas bubbles and their rate of rise

Insight on the investigated impact of heat exchanging internals on the hydrodynamics of bubble columns Given

Problem definition

A scale-up methodology is proposed based on controlling the effect of scale using internals by means of reactor compartmentalization

The logical final step is to replicate the single bundle and to re-investigate the hydrodynamics within each, versus the obtained in the 6” diameter column.

Reactor compartmentalization approach

Basis and key questions

Hypothesis

How to accomplish this?The large reactor diameter is subdivided into

similar, vertical compartments by means of the cooling tubes.

The compartments are to have a diameter similar to that of a small scale column on which investigations can be performed.

The various hydrodynamic parameters within each compartment are to be compared with those measured in a bubble column of the same diameter.Experimental investigation: Design and procedure

Schematic diagram of empty 18” column

6”

Step 1: compare the hydrodynamics generated inside the 6” tubes bundle with the corresponding results in the 6” bubble column.

“Overall gas holdup and its radial profile or cross-sectional distribution should be the same for two reactors to be dynamically similar” - Shaikh, 2007

1 central bundle-19 PVC rods covering 5% of the total CSA.

Preliminary results

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

20 cm/s

30 cm/s

45 cm/s

Dimensionless radius (r/R)

Gas h

oldu

p

Radial gas holdup profiles inside the tube circle bundle exhibited similar behavior as a solid wall column being almost parabolic at the investigated high gas velocities

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 10

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

20 cm/s

r/R=0.3 from empty 18 inch

(r/R) - For the 6" bundle

Gas H

oldu

p

(r/R) - For the empty 18" column

The tubes showed an effect close to that of a solid wall (low local εg value) when comparing gas holdup values at r/R = 0.9 inside the circular tube bundle with the equivalent dimensionless radius in the empty column.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10.000.050.100.150.200.250.300.350.400.450.500.55

30 cm/s (Junli Xue in empty 6")30 cm/s

Dimensionless radius (r/R)

Gas h

oldu

p

Next step

Can vertical internals be arranged to form small internal bubble column within the large reactor wall?

Can vertical internals, in such arrangement, mimic the behavior of columns of the same smaller diameter having a solid wall instead?

A close agreement

between resulting gas holdup profile inside the single tube bundle and

the data in 6” bubble column is

observed.

This scale-up design is sought to offer the following advantages: A scale-up methodology considering the exothermic nature of reactions occurring in bubble columns and the associated cooling tubes. A simple, yet efficient, way of designing large scale bubble column reactors via reactor compartmentalization into small size columns. Better prediction of the performance of bubble columns based on a better control on the effect of scale. A substantial decrease in phase backmixing usually observed in large diameter columns. A much uniform bubble size.Acknowledgments

Ahmed Youssef, Milorad Duduković, Muthanna Al-Dahhan

AY 08

(GTL.F1)