Fluidization, solids handling, and processing: industrial applications

895 Pages·1998·12.356 MB·English

by  Wen-ching Yang

Preface
This volume, Fluidization, Solids Handling, and Processing, is
the first of a series of volumes on "Particle Technology" to be published
by Noyes Publications with L. .S Fan of Ohio State University as the
consulting editor. Particles are important products of chemical process
industries spanning the basaincd specialty chemicals, agricultuprraold ucts,
pharmaceuticals, paints, dyestuffs and pigments, cement, ceramics, and
electronic materials. Solids handling and processing technologies are
thus essential to the operation and competitiveness of these industries.
technology Fluidization is employed not only in chemical production, it also
is applied in coal gasification and combustion for power generation, mineral
processing,pf rooocde ssing, soil washing and other related waste treatment,
environmental remediation, and resource recovery processes. The FCC
(Fluid Catalytic Cracking) technology commonly employed in the modem
petroleum refineries is also based on the fluidization principles.
There are already many books published on the subjects offluidiza-
tion, solids handling, andprocessing. On first thought, I was skepticaablo ut
the wisdom and necessity of one more book on these subjects. On closer
examination, however, I found that some industrially important subjects
were either not covered in those books or were skimpily rendered. It would
be a good service to the profession and the engineering community to
assemble all these topics in one volume. In this book, I have invited
recognized experts in their respective areas to provide a detailed treatment
vi
Preface 7lv
of those industrially important subjects. The subject areas covered in this
book were selected based on two criteria: )i( the subjects are of industrial
importance, and (ii) the subjects have not been covered extensively in
books published to date.
The chapter on fluidized bed scaleup provides a stimulating
approach to scale up fluidized beds. Although the scaleup issues are by no
means resolved, the discussion improves the understanding of the issues and
provides reassessments of current approaches. The pressure and tem-
perature effects and heat transfer ni fluidized beds are covered in
separate chapters. They provide important information to quantify the
effects of pressure and temperature. The gas distributor and plenum
design, critical and always neglected in other books, are discussed in detail.
For some applications, the conventional fluidized beds are notnecessarily the
best. Special designf eatures can usuallayc hietvhee objective cheaper and
be more forgiving. Two of the non-conventional fluidized beds, recirculat-
ing fluidized beds with a draft tube and jetting fluidized beds, are intro-
duced and their design approaches discussed. Fluidized bed coating and
granulation, applied primarily in the pharmaceutical industry, is treated
from the fluidization and chemical engineering point of view. Attrition,
which is critical in design and operation of fluidizedb eds and pneumatic
transport lines, is discussed in detail in a separate chapter. Fluidization with
no bubbles to minimize bypassing, bubbleless fluidization, points to
potential areas of application of this technology. The industrial applica-
tions of the ever-increasingly important three-phasefluidization systems
are included as well. The developments in dense phase conveying and in
long distance pneumatic transport with pipe branching are treated sepa-
rately in two chapters. The cyclone, the most common component em-
ployed in plants handling solids and often misunderstood, is elucidated by
an experiencedp ractitioner in the industry. The book is concluded with a
discussion on electrostatics and dust explosion by an electrical engineer.
This book is not supposed to be all things to all engineers. The
primary emphasis of the book is for industrial applications atnhde primary
audience is expected to be the practitioners of the art of fluidization, solids
handling, and processing. It will be particularly beneficial for engineers who
operate or design plants where solids are handled, transported, and pro-
cessuesdi ng fluidization technology. The book, however, can also be useful
as a reference book for students, teachers, and managers who study particle
technology, especially in the areas of application offluidization technology
and pneumatic transport.
viii Preface
I'd like to take this opportunity to thank Professor Fan who showed
confidence in me to take up this task and was always supportive. I'd also
like to thank the authors who contributed to this book despite their busy
schedules. All of them are recognized and respected experts in the areas
they wrote about. The most appreciation goes to my wife, Rae, who
endured many missing weekends while I worked alone in the office,
Pittsburgh, Pennsylvania Wen-Ching Yang
February, 1998
NOTICE
To the best of our knowledge the information in this publication is
accurate; however the Publisher does not assume any responsibility
or liability for the accuracy or completeness of, or consequences
from, arising such for intended is book information. This informational
purposes only. Mention of trade names or commercial products does
not use for recommendation or endorsement constitute by Publisher. the
Final determination of the suitability of any information or product
for use contemplated by any user, andt he manner of that use, is the
sole responsibility of the user. We recommend that anyone intending
to rely on any recommendation of materials or procedures mentioned
in this publication should satisfy himself as to such suitability, and
that he can meet all applicable safety and health standards.
Contributors
John C. Chen Thomas B. Jones
Department of Chemical Department of Electrical
Engineering Engineering
Lehigh University University of Rochester
Bethlehem, PA Rochester, NY
Bryan J. Ennis S.B. Reddy Karri
E&G Associates Particulate Solid Research, Inc.
Nashville, TN Chicago, IL
Liang-Shih Fan George E. Klinzing
Department of Chemical Department of Chemical and
Engineering Petroleum Engineering
Ohio State University University of Pittsburgh
Columbus, OH Pittsburgh, PA
Leon R. Glicksman Ted M. Knowlton
Department of Architecture, Particulate Solid Research, Inc.
Building Technology Program Chicago, IL
Massachusetts Institute of
Technology Mooson Kwauk
Cambridge, MA Institute of Chemical Metallurgy
Adacemia Sinica
Beijing, People's Republic of
China
x/
x Contributors
Jack Reese Joachim Werther
Department of Chemical Hamburg- University Technical
gnireenignE grubraH
Ohio State University Germany Hamburg,
Columbus, HO
Peter Wypyeh
Jens Reppenhagen Department of Mechanical
University Hamburg- Technical gnireenignE
grubraH University of Wollongong
Germany Hamburg, Wollongong, NSW, Australia
Ellen M. Silva Shang-Tian Yang
Department of Chemical Department of Chemical
gnireenignE gnireenignE
Ohio State University Ohio State University
Columbus, HO Columbus, HO
Gabriel I. Tardos Wen-Ching Yang
Department of Chemical Science dna Technology Center
gnireenignE Power Westinghouse Siemens
College City of University City of noitaroproC
New York PA Pittsburgh,
New York, NY
Frederick A. Zenz
Richard Turton Modeling & Equipment Process
Department of Chemical Mfg. Co., Inc.
gnireenignE Cold Spring, NY
University Virginia West
Morgantown, WV
1
Fluidized Bed Scale-up
Leon R. Glicksman
1.0 INTRODUCTION
Although fluidized beds have been used extensively in commer-
cial operations such as fluidized bed combustors and fluid catalytic crack-
ing, engineers are still faced with uncertainties when developing new
commercial designs. Typically, the development process involves a
laboratory bench scale unit, a larger pilot plant, and a still larger demon-
stration unit. Many of the important operating characteristics can change
between the different size units. There is a critical problem of scale-up:
how to accurately account for the performance changes with plant size to
insure that a full size commercial unit will achieve satisfactory perfor-
mance. In addition, it would be helpful if the smaller units could be used
to optimize the commercial plant or solve existing problems.
One discouraging problem is the decrease in reactor or combustor
performance when a pilot plant is scaled up to a larger commercial plant.
These problems can be related to poor gas flow patterns, undesirable solid
mixing patterns and physical operating problems (Matsen, 1985). In the
synthol CFB reactors constructed in South Africa, first scale-up from the
pilot plant increased the gas throughput by a factor of 500. Shingles and
McDonald (1988) describe the severe problems initially encountered and
their resolution.
Fluidization, Handling, Solids and Processing
In some scaled up fluidized bed combustors, the lower combus-
tion zone has been divided into two separate subsections, sometimes
referred to as a "pant leg" design, to provide better mixing of fuel and
sorbent in a smaller effective cross section and reduce the potential
maldistribution problems in the scaled up plant.
Matsen (1985) pointed out a number of additional problem areas
in scale-up such as consideration of particle size balances which change
over time due to reaction, attrition and agglomeration. Erosion of cy-
clones, slide valves and other components due to abrasive particles are
important design considerations for commercial units which may not be
resolved in pilot plants.
If mixing rates and gas-solid contacting efficiencies are kept
constant between beds of different size, then thermal characteristics and
chemical reaction rates should be similar. However, in general, the bed
hydrodynamics will not remain similar. In some instances, the flow
regime may change between small and large beds even when using the
same particles, superficial gas velocity and particle circulation rate per
unit area. The issue of scale-up involves an understanding of these
hydrodynamic changes and how they, in turn, influence chemical and
thermaclo nditions by variations in gas-solid contact, residence time, solid
circulation and mixing and gas distribution.
There are several avenues open to deal with scale-up. Numerical
models have been developed based on fundamental principals. The
models range from simple one-dimensional calculations to complex mul-
tidimensional computational fluid dynamics solutions. There is no doubt
thats uch first principal models are a great aid in synthesizintge st data and
guiding the development of rational correlations. In a recent model
evaluation, modelers where given the geometry and operating parameters
for several different circulating beds and asked to predict the hydrody-
namic characteristics without prior knowledge of the test results (Knowlton
et al. 1995). None of the analytical or numerical models could reliably
predict all of the test conditions. Few of the models could come close to
predicting the correct vertical distribution of solid density in the riser and
none could do it for all of the test cases! Although it is tempting to think
that these problems can be solved with the "next generation of comput-
ers," until there is general agreement and thorough verification of the
fundamental equations used to describe the hydrodynamics, the numerical
models will not stand alone as reliable scale-up tools.
On the other hand, there is a blizzard of empirical and semi-
empirical correlationwsh ich exist in the fluidized bed literature to predict
Fluidized Bed Scale-up 3
fluid dynamic behavior. In addition there are probably a large number of
proprietary correlations used by individual companies. The danger lies in
extrapolating these relations to new geometric configurations of the riser
or inlet, to flow conditions outside the range of previous data, otro beds of
much different sizes. Avidan and coauthors in a 1990 review of FCC
summed up the state of the art: "basic understanding of complex fluidiza-
tion phenomena is almost completely lacking. While many FCC licensors
and operators have a large body of in-house proprietary data and correla-
tions, some of these are not adequate, and fail when extrapolated beyond
their data base." (Avidan, et al., 1990.)
As a example, consider the influence of mean particle size. In the
early work on bubbling fluidized bed combustors, attempts were made to
use relations from the classic fluidization literature which had concen-
trated on FCC applications with much smaller particles. In many cases, it
was discovered that the relationships for small particles gave erroneous
results for combustors with much larger particles. For example, the two phase
theory equating the excess gas velocity above minimum fluidization to the
visible bubble flow was shown to be severely distorted for large particle
systems. Jones and Glicksman (1985) showed that the visible bubble flow
in a bubbling bed combustor was less than one fifth of .fmU-ou In other
cases even the trends of the parametric behavior were changed. Heat transfer
to immersed surfaces in fine particle bubbling beds increased strongly
with a decrease in the mean particle size. For large particle beds, the heat
transfer, in some instances, decreased with a decreased particle diameter.
Another approach to scale-up is the use of simplified models with
key parameters or lumped coefficients found by experiments in large
beds. For example, May (1959) used a large scale cold reactor model
during the scale-up of the fluid hydroforming process. When using the
large cold models, one must be sure that the cold model properly simulates
the hydrodynamics of the real process which operates at elevated pressure
and temperature.
Johnsson, Grace and Graham (1987) have shown one example of
verification of a model for 2.13 m diameter industrial phthalic anhydride
reactor. Several bubbling bed models gave good overall prediction of
conversion and selectivity when proper reaction kinetics were used along
with a good estimate of the bubble size. The results were shown to be
quite sensitive to the bubble diameter. The comparison is a good check of
the models but the models are incomplete without the key hydrodynamic
data. In this case, the bubble size estimates were obtained from measure-
ments of overall bed density in the reactor.
Fluidization, Solids Handling, and Processing
As Matsen expresses it, after over a half a century of scale-up
activity in the chemical process industry, "such scale-up is still not an
exact science but is rather a mix of physics, mathematics, witchcraft,
history and common sense which we call engineering." (Matsen, 1995.)
A complete treatment of scale-up should include the models,
numerical calculation procedures ande xperimental data designers need to
carry out successful scale-up from small size beds to commercial units.
This would involve a large measure of the existing fluidized bed research
and development effort; clearly, such a task is beyond the scope of a single
chapter. Since changes in the bed size primarily influence scale-up
throughc hanges in the bedh ydrodynamics, one focus of this chapter is on
experimental results and models which deal explicitly with the influence
of bed diameter on hydrodynamic performance for both bubbling and
circulating fluidized beds. The changes in the bed dynamics will, in turn,
impact the overall chemical conversion or combustion efficiency through
changes in the particle-to-gas mass transfer and the heat transfer from the
bed to immersed surfaces or the bed wall. Several examples of this
influence are also reviewed.
The second focus of this chapter is on the use of small scale
experimental models which permit the direct simulation of the hydrody-
namics of a hot, possibly pressurized, pilotp lant or commercial bed. By
use of this modeling technique, beds of different diameters, as well as
different geometries and operating conditions, can be simulated in the
laboratory. To date, this technique has been successfully applied to
fluidized bed combustors and gasifiers. Derivation of the scale modeling
rules is presented for a variety of situations for gas solid fluidized beds.
Verification experiments and comparisons to large scale commercial
systems are shown. Rules for the use of this experimental modeling
technique for FCC operations as well as fotrh e simulation of bed-to-solid
surface heat transfer are also given.
2.0 REACTOR MODELING: BED DIAMETER INFLUENCE
In this section, representative results are reviewed to provide a
prospective of reactor modeling techniques which deal with bed size.
There probably is additional unpublished proprietary material in this area.
Early studies of fluidized reactors recognized the influence of bed diam-
eter on conversion due to less efficiengta s-solid contacting. Experimental
studieswere used to predict reactor performance. Frye et al. (1958) used
Fluidized Bed Scale-up 5
a substitute reaction of ozone decomposition to study hydrocarbon synthe-
sis. The ozone decomposition can be run at low pressures and tempera-
tures and can be rate-controlled in the same way and by the same catalyst
as the reaction under development. Frye and coworkers used three beds of
2 inch, 8 inch and 30 inch diameter, respectively, to study the size
influence. We should interject a caution that the use of pressures and
temperatures different than the actual reaction may mean that the hydro-
dynamics of the substitute reaction model will differ from the actual
application; this is illustrated later in the chapter. Figure 1 shows the
apparent reaction rate constant for the different bed diameters at two
different bed heights with the other parameters held constant. Note that
the rate constant decreased by roughly a factor of three between the 2 inch
and 30 inch beds.
lx10" -~
!
8x10 s
6xlO'S !
4x 10 5"
0.$ .:I"11 / $s dmO (cid:12)9
i~dcr~t s I,tts sY Ut( dS I'll. I, (cid:12)9 ItEdCll~t
AnO mLs ROF EJlC# Old. FO
z Cdlrd/,IrST I~#lrlc&s StZs IrO#s ORIIIO
< 2xlO-S
I-
z
o
to
-M-,II
:1< lx10.5 __..~....
8xlO 6"
O Z (cid:12)9
_._..,i.,,~
r~ 6xl (14
uJ
4x10 6"
Z
ul
<
2x 104
1 2 4 6 8 10 20 40 60 80 100
REACTOR DIAMETER, inches
Figure .1 Apparent etar-noitcaer tnatsnoc .sv diameter reactor dna height. bed (From
Frye et al., 1958.)