Group Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Title: 14.5.3 - Investigation of the distributed addition of gas to a fluidized bed reactor
ALL VOLUMES CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00102023/00357
 Material Information
Title: 14.5.3 - Investigation of the distributed addition of gas to a fluidized bed reactor Fluidized and Circulating Fluidized Beds
Series Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Physical Description: Conference Papers
Creator: Lemin, B.
Heinrich, S.
Hartge, E.-U.
Publisher: International Conference on Multiphase Flow (ICMF)
Publication Date: June 4, 2010
 Subjects
Subject: secondary air addition
fluidized bed membrane reactor
fluid mechanics
 Notes
Abstract: A concept to increase yield and selectivity during the partial oxidation of hydrocarbons is the distributed addition of oxygen. In the current paper the influence of the distributed addition of gas to a fluidized bed reactor on the fluid mechanics has been studied. The focus of this contribution is to study two different concepts of dosage of reactants by using vertical and horizontal porous tubes in a fluidized bed. The analysis of the fluid dynamics has been carried out under cold and hot gas atmospheres, with and without the tube membranes. A fiber optical sensor for cold gas conditions has been used to measure solids concentrations and velocities of the solid catalysts in the fluidized bed. Additionally Particle Image Velocimetry was used to determine the flow field. To study the penetration and distribution of the gas fed via the porous tubes a gas potentiometric oxygen sensor for hot gas conditions was used. The results of the different measurements show, that fluidization is more even and less vigorous when a fraction of the total gas is injected distributed along the height. The large scale circulation and thus also the back mixing is significantly reduced. Comparing the vertical and the horizontal tubes, the results indicate that most of the gas from the horizontal tubes forms goes directly into the bubble phase, while for the vertical tubes the major part of the gas is released into the suspension phase, which should be preferable for the heterogeneously catalyzed partial oxidations.
General Note: The International Conference on Multiphase Flow (ICMF) first was held in Tsukuba, Japan in 1991 and the second ICMF took place in Kyoto, Japan in 1995. During this conference, it was decided to establish an International Governing Board which oversees the major aspects of the conference and makes decisions about future conference locations. Due to the great importance of the field, it was furthermore decided to hold the conference every three years successively in Asia including Australia, Europe including Africa, Russia and the Near East and America. Hence, ICMF 1998 was held in Lyon, France, ICMF 2001 in New Orleans, USA, ICMF 2004 in Yokohama, Japan, and ICMF 2007 in Leipzig, Germany. ICMF-2010 is devoted to all aspects of Multiphase Flow. Researchers from all over the world gathered in order to introduce their recent advances in the field and thereby promote the exchange of new ideas, results and techniques. The conference is a key event in Multiphase Flow and supports the advancement of science in this very important field. The major research topics relevant for the conference are as follows: Bio-Fluid Dynamics; Boiling; Bubbly Flows; Cavitation; Colloidal and Suspension Dynamics; Collision, Agglomeration and Breakup; Computational Techniques for Multiphase Flows; Droplet Flows; Environmental and Geophysical Flows; Experimental Methods for Multiphase Flows; Fluidized and Circulating Fluidized Beds; Fluid Structure Interactions; Granular Media; Industrial Applications; Instabilities; Interfacial Flows; Micro and Nano-Scale Multiphase Flows; Microgravity in Two-Phase Flow; Multiphase Flows with Heat and Mass Transfer; Non-Newtonian Multiphase Flows; Particle-Laden Flows; Particle, Bubble and Drop Dynamics; Reactive Multiphase Flows
 Record Information
Bibliographic ID: UF00102023
Volume ID: VID00357
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: 1453-Lemin-ICMF2010.pdf

Full Text


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Investigation of the distributed addition of gas to a fluidized bed reactor


Bert Lemin1,2, Stefan Heinrichl, Ernst-Ulrich Hartgel

SHamburg University of Technology, Institute of Solids Process Engineering and Particle Technology
Denickestrasse 15, D-21073 Hamburg, Germany
SOtto-VOn-Guericke-University Magdeburg, Institute of Process Equipment and Environmental Technology
Universitiitsplatz 2, D-39106 Magdeburg, Germany
hartge~tuhh.de


Keywords: secondary air addition, fluidized bed membrane reactor, fluid mechanics




Abstract

A concept to increase yield and selectivity during the partial oxidation of hydrocarbons is the distributed addition of oxygen. In
the current paper the influence of the distributed addition of gas to a fluidized bed reactor on the fluid mechanics has been
studied. The focus of this contribution is to study two different concepts of dosage of reactants by using vertical and horizontal
porous tubes in a fluidized bed. The analysis of the fluid dynamics has been carried out under cold and hot gas atmospheres,
with and without the tube membranes. A fiber optical sensor for cold gas conditions has been used to measure solids
concentrations and velocities of the solid catalysts in the fluidized bed. Additionally Particle Image Velocimetry was used to
determine the flow field. To study the penetration and distribution of the gas fed via the porous tubes a gas potentiometric
oxygen sensor for hot gas conditions was used. The results of the different measurements show, that fluidization is more even
and less vigorous when a fraction of the total gas is injected distributed along the height. The large scale circulation and thus
also the back mixing is significantly reduced. Comparing the vertical and the horizontal tubes, the results indicate that most of
the gas from the horizontal tubes forms goes directly into the bubble phase, while for the vertical tubes the major part of the
gas is released into the suspension phase, which should be preferable for the heterogeneously catalyzed partial oxidations.


superficial fluidizing velocity (m/s)
solids velocity (m/s)


Introduction


Subscripts
mf minimum fluidization condition


Experimental

Three different setups have been used for the experiments,
the first two being a cold model (Figure 1) made of resin.
The second unit for high temperature operation was made
from stainless steel (Figure 2). Both units have a
cross-section of 100x100 mm2 and a total height of 300 mm.
The latter unit could be electrically heated to temperatures
of up to 6000C.
The resin fluidized bed has been equipped with horizontal
measuring ports in 9 heights. In each height 2 or alternating
3 ports have been provided (Figure 1). The ports were
located on the side wall, such that probes were inserted
rectangular to the horizontal membrane tubes. The hot unit
was equipped with a measuring port in the top plate of the
unit, located 4 cm from the front and 4 cm from the right
hand wall. Thus a probe inserted through this port measured
on a line in a distance of about 5 mm from the middle
horizontal membrane tubes and about the same distance to
the central vertical membrane tube.


The partial oxidation and the oxidative dehydrogenation of
hydrocarbons are important reaction types in the industrial
organic chemistry (Weissermehl and Arpe, 2003). Both
types of reactions are heterogeneously catalyzed gas phase
reactions which are often carried out in fluidized bed
reactors. Safety and process economy are pre-conditions to
handle these thermal chemical processes. The separated
dosage of reactants is a possibility with high potential to
increase yield and selectivity and also safety of operation.
Ahchieva et al. (2I III and Deshmukh et al. (2I II) have
demonstrated for the example of the catalytic oxidative
dehydrogenation of ethane that the distributed addition of
the air distributed along the height of the fluidized bed
further increases process efficiency.
The focus of this contribution is to evaluate two different
concepts of distributed dosage of reactants by using porous
tubes which are arranged horizontally or vertically in the
fluidized bed reactor. The influence of these dosing
concepts on fluid dynamics and reactant distribution was
studied experimentally in this work.

Nomenclature

c, Solids volume concentration (-)
S/P Secondary to primary air ratio ( )






7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Figure 1: Resin test unit with Figure 2: High temperature
ports probe ports and vertical unit
membranes installed

Figures 3 and 4 show images of the porous tubes used for
the gas distribution. Figure 3 shows the tube package with
the horizontal tubes. It consists of 3 rows with 3 tubes in the
lowest, 2 tubes in the middle and 3 tubes in the upper row.
For the gas supply the tubes penetrate through the rear wall,
outside of the bed they are then connected to a gas
distributor. Figure 4 shows the packet with the vertically
installed tubes. The packet is installed into the fluidized bed
from the top, gas is also supplied to the membrane tubes
from the top. This bundle consists of 3 x 3 rows in a
rectangular arrangement.
The porous tubes for gas distribution were made of sintered
metal and had an inner diameter of 3 mm, an outer diameter
of 6 mm and a pore width of about 8 Clm.
The third setup was a two-dimensional unit with a
cross-section of 100x10 mm2 and a height of 400 mm. This
unit was used for measurements by particle image
velocymetrie The arrangement of the horizontal tubes was
the same as for the other two units with the tubes in
direction of the short side. For the vertical arrangement of
the tubes one row of three tubes was installed in the mid
between the front and rear windows.

Measurement Techniques

Different in situ measurement techniques were used for the
investigation of fluid dynamics. A fiber optical sensor (FOS)
(Figure 5) was used to measure velocities and solids
concentrations in the
fluidized bed under ambient 6


Figure 5: Two channel fiber optical sensor, d=6/2 mm


idizdbfd


DC Lamps


CMOS Highspeed Camera








Computer


200mm
losomm


Figure 6: Setup used for particle image velocimetry
conditions. Velocities are determined by cross correlation of
the two channels of the probe. 300 k samples were taken
each time with a frequency of 10 k To get further information on the influence of the different
gas distribution concepts on the solids movement in the
fluidized bed particle image velocimetry (PIV) was utilized.
The setup for these measurements is depicted in Figure 6.
Investigation of back mixing was conducted under hot
conditions with temperatures of nearly 6000C. A gas
potentiometric oxygen probe (GOP) (Figure 7) allows fast
detection of instantaneous oxygen concentrations. For
details of the GOP measurement see Schotte et al. (2010).
For these measurements 10 k samples were taken at a
frequency of 1 kHz.


Figure 7: Gas potentiometric oxygen probe (GOP),
d= 12/6 mm, multi-sensor


Results and Discussion

Measurement of velocity and solid concentration

Measurements with the fiber optical probe have been made
at 23 ports at 5 different lateral positions. Measurements in
a plane close to the wall (1 cm from the wall) and in the
central plane (5 cm distance from the wall) are shown for
measurements without tubes and with horizontal tubes in
Figure 8 and Figure 9, respectively.
The contour plots generated by interpolation from the local
measurements show the distribution of solid concentrations.


Figure 3: Top view on the
horizontal membranes
installed into the hot unit.


Figure 4: Bundle of
vertically membranes.






7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


S/P = 0 S/P = 1


10,60
18- 18 c

16 16 0,50

14 14 0,40

O 12- 12
E 0,30
.9 10 10

8- 1 81 0,20


4-1 4 I0,10

2 2 0,00
-2 -1 0 1 2 -2 -1 0 1 2
y, cm y, cm
x =1cm x = 5cm
Figure 8 Solids volume concentration c,, without tubes.
Ut= 5 -i,,,

The total fluidization gas was adjusted to ut= 5 t,. The
plot in Figure 8 has been derived from measurements
without any tubes inside the bed, while Figure 9 shows
measurements with horizontal tubes and a split of the total
gas into equal flows of primary and secondary air, thus
giving a secondary to primary air ratio S/P = 1/1.
The measurements with primary air only (Figure 8) show
distinct profiles of the solids volume concentration in the
bed, with lower concentrations in the center and higher
concentrations near the wall. This indicates a movement of
the bubbles into the core region of the bed. The bubbles then
explode at the surface and throw particles into the freeboard
which will fall down near the wall, causing the elevated
solids concentration near the wall in the lower freeboard
region. With the split of the fluidizing gas into primary and
secondary air (Figure 9), the solids distribution become
much more even, indicating that also the bubbles are
distributed more homogeneously. The bed height is slightly
decreased for the second case. The reason for this is the



MM 10,60
18 18 C,

16 16 oI o


1,00
v, m/s
0,80
0,60
0,40
0,20
0,00
-0,20
-0,40
-0,60
-0,80
-1,00


2 &
-2-1 0 1 2
y, cm
x=5cm


-2 -1 0 1 2
y, cm
x=5cm


Figure 10: Velocity measurements with the fiber optical
sensor, left: S/P= 0, right: horizontal tubes, S/P=1/1

smaller bubble hold-up in the bed with distributed gas
addition, which results from the smaller gas volume passing
the lower part of the bed. The full gas volume passes only
the upper part above the uppermost gas addition, Therefore
the average gas volume flow is significantly smaller than
with primary air only.
With the fiber optical sensor also rise velocities of flows
structures, i.e. mainly bubbles in the bed, have been
determined. The shown in Figure 10 shows as an example
the comparison of the velocities measured in the center
plane of the bed with horizontal tubes installed. The plot on
the left is for operation without secondary gas addition, the
plot on the right for operation with secondary air (S/P = 1).
The comparison again indicates that for distributed gas
dosage bubbles are smaller and thus slower and show less
tendencies to move towards the center of the unit.

PlV measurements

In Figure 11 results of the PIV-measurements are shown.
Measurements without any inserted tubes are shown in the
left column. The second and third columns show
measurements with the horizontal tubes inserted, the second
column without any secondary air flow, the third column
with two third of the air entering as primary air and one
third as secondary air (S/P = 1/2). The left two columns
show the results gained with vertical tubes installed, with
and without secondary air, respectively. The upper row in
Figure 11 gives the value of the velocity component in
vertical direction, the bottom row streamlines of the solids
flOW.
Looking at the upper row, the PIV-measurements are in
good agreement with the results from the probe
measurements with the fiber optical sensor. The
measurements without any tubes show high upward
velocities in the center, indicating bubbles moving up
preferably in the center. Installing tubes either horizontally
or vertically slightly reduce this concentration of the
bubbles in the center, but do not change the general picture.
But when the fluidization gas is split into primary and
secondary air, the picture completely changes: There is no
significant maldistribution to be seen any more.


14-v v

12-

lo-

8 -

6-

4-

2-
-2 -1 0 1 2
y, cm
x = 5cm


0,40


0,30


0,20





S 010,00


-2 -1 0 1 2
y, cm
x = cm


Figure 9: Solids volume concentration c,, horizontal tubes,
S/P =1, it= 5 i






















































20
15C- 5/P=1/3
~P -5/P=3/1





0 2 4 6 8 10 12 14 16



*-* 1,5 -

S1 - - - -- -

0 .5 - - -- - - -


0 2 4 6 8 10 12 14 16


20

15


6,


O 2 4 6 8 10 12 14 1




1 -





O 2 4 6 8 10 12 14 1


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


without
tubes


horizontal tubes
S/P = 0 S/F


vertical tubes
P = 1/2 S/P = 0 S/P = 1/2




'11)


1200 I0.6
I 2000.4

0.2


0 E


-0.4




-0.6
S0.6

0.2

0 E

-0.2



1 -0.4-.


100




E
0 E

200 .2




100


0 25 50 75 100 0 25 50 75 100 0 25 50 75 10
width, mm

Figure 11: Results of PIV-measurements. Top row: velocity
colored with velocity value. u = 7 u f, d, = 600 Clm

The bottom row of Figure 11 shows the streamlines of the
solids flow. For the case without tubes in the bed a gross
circulation of the solids can be noted. This gross circulation
is induced by the upward movement of the bubbles in the
center region. It is significantly reduced by the distributed
addition of secondary air. A slight reduction can already be
recognized with the installed tubes only without any


0 25 50 75 100| 0 25 50 75 100


component in vertical direction, bottom row: streamlines,


gas penetration into the bed. For the measurement of the gas
penetration Nitrogen was used as primary gas and air as
secondary gas. For the measurements of the gas
concentrations the GOP probe was inserted vertically
through the port in the top of the unit.
Figure 12 shows the results for the gas mixing
measurements. The upper two figures show the measured


secondary air addition.
Here one possible risk of the
distributed addition of secondary air
becomes obvious: since the gross
circulation causes large scale
movement of the solids, suppressing
the circulation is equivalent to
hindering this wide distance mixing.
In general solids mixing and also
heat homogeneity therefore might be
worse for fluidized beds with
distributed air addition.

Measurement of the gas
penetration
Beside the measurements of the pure
fluid mechanics as described before,
additional measurements under hot
conditions with the gas
potentiometric oxygen probe have
been carried out. These
measurements give additional
information on gas back mixing and


Figure 12: Vertical profiles of the oxygen concentration and the standard deviation SD
of the oxygen concentration measured by the GOP probe. u = 6.7 um


11






7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

dynamics. It reduces bubble size and results in a more
homogeneous fluidization throughout the cross-section of
the bed. On the other hand the gross mixing will be
hindered. With respect to the fluid mechanics no significant
difference between horizontal and vertical membrane tubes
could be observed. But the penetration of the gas into the
bed differs significantly between the horizontal tube
membranes and the vertical ones. The instantaneous
measurements of the oxygen concentration with the GOP
probe indicate that the secondary gas is fed with the vertical
membranes more evenly distributed into both phases, the
suspension and the bubble phase, whereas at the horizontal
tubes new bubbles are formed by the secondary gas and
therefore the major fraction of the secondary gas is flowing
into the bubble phase. With respect to the partial oxidations
the injection of the secondary gas is preferred, since in the
suspension phase the catalyzed partial oxidation will take
place while in the solids free bubble phase already formed
product gas might be further oxidized with the fresh
oxygen.

Acknowledgements

This work was supported by the Deutsche
Forschungsgemeinschaft within the proj ect HE 4526/1 -4.

References

Ahchieva, D., B. Lemin, S. Heinrich, and L. M~rl c(2 .~
Controlled addition of reactants in a fluidized bed
membrane reactor. In J. W. W. N. K.-E. W E.-U. Hartge
(Ed.), Circulating Fluidized Bed Technology IX, Hamburg,
pp. 421-426. TuTech Innovation GmbH.

Deshmukh, S., S. Heinrich, L. M~rl, M. Sint Annaland van,
and J.Kuipers IalIIs=). Membrane assisted fluidized bed
reactors: Potentials and hurdles. Chemical Engineering
Science 62(1-2), 416-436.

Link, J., W. Godlieb, P. Tripp, N. Deen, S. Heinrich,
J. Kuipers, M. Schinherr, and M. Peglow .
Comparison of fibre optical measurements and discrete
element simulations for the study of granulation in a spout
fluidised bed,. Powder Technol. 189, 202-217.

Schotte, E., B. Lemin, H. Lorenz, and H. Rau (2010). Gas
potentiometry. In M. Lackner, F. Winter, and A. K. Agarwal
(Eds.), Handbook of Combustion, Volume Vol. 2:
Combustion Diagnostics and Pollutants. Willey-VCH.

Weissermehl, K. and H.-J. Arpe :<2 1I). Industrial Organic
Industry (4th. rev. ed. ed.). Weinheim: Wiley VCH.


oxygen concentration, on the left for the vertically arranged
membrane tubes, on the right for the horizontal tubes. The
pictures in the second row show the standard deviation of
the measurements. The positions of the membrane tubes are
indicated in the second row of diagrams. The expanded bed
height was about 16 cm, thus there are no measurements in
the freeboard shown. With the GOP no measurements
without secondary gas feeding are possible. Since there is
no reaction taking place a separated feeding of the oxygen
containing and the non oxygen containing gases is
necessary to produce any uneven distribution of the oxygen.
Feeding both gases through the gas distributor would give a
constant oxygen concentration throughout the whole bed.
From theory oxygen concentration should be zero upstream
of the membranes. For the vertical membrane tubes a linear
increase of the Oz concentration can be expected starting at
the lower end of the membrane tube. For the horizontal
tubes a stepwise increase would be reasonable. The end
concentration should be proportional to the ratio S/P of
secondary and primary gas.
For the measurement with the low amount of secondary gas
S/P=1/3 some oxygen has been measured below the
membrane, indicating some back mixing. It is not clear, if
the oxygen is transported by real gas back mixing or if it is
adsorbed to the porous solids and then transported with the
back mixed solids. For the higher secondary air mass flows
and therefore lower velocities in the region upstream of the
membrane tubes no indication of back mixing of oxygen
can be found in the concentration plots. Only the plots of
the standard deviation SD indicate also for the ratio S/P=1/1
some minor back mixing. As expected the concentration
plots for the vertical tubes show a nearly linear increase of
the Oz concentration. But it starts about 2 cm downstream of
the lower end of the tube. This is mainly due to the limited
horizontal mixing and the distance of about 0.5 cm between
the GOP probe and the membrane tube. During the time
needed to transport the gas by dispersion 5 mm in horizontal
direction, it is transported about 20 mm by convection in
vertical direction. For the horizontal tubes no stepwise
increase of the concentration can be recognized. Obviously
the axial mixing due to the bubble behavior is strong
enough to level out the local addition of the secondary air.
From the standard deviation which is much higher for the
horizontal tubes than for the vertical tubes it can bed
deduced, that at the horizontal tubes more bubbles are
formed by the secondary gas than with the vertical tubes. In
that case the GOP probe will be measure alternating in the
bubble phase with its high Oz concentration and then in the
suspension phase with the low oxygen concentration. This
strongly alternating signal will the give a high value for the
standard deviation SD. For the vertical tubes obviously
more secondary gas is flowing directly into the suspension
phase, thus lowering the concentration difference between
bubble and suspension phase and therefore lowering the
standard deviation SD. With increasing height in both cases
the concentration difference between bubble and suspension
phase levels of by the gas exchange between bubbles and
suspension.

Conclusions

Distributed addition of secondary gas by porous membranes
in fluidized bed reactors significantly influences the fluid




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs