Você está na página 1de 13

Pressure Drop in Vertical Tubes

in Transport of Solids by Gases


can be regarded as t h e sum of t h e drop due t o t h e carrier
gas alone plus a solids pressure drop. T h e residual obtained by subtracting t h e solids static head from t h e total
solids pressure drop was treated as an apparent solids
friction drop, which conformed t o t h e Fanning equation
when solids velocity and dispersed solids density were
used. However, later measurements showed t h a t t h e
m a j o r portion of t h e apparent friction was due t o particle
acceleration in t h e test section. T h e mathematical treatm e n t of t h e accelerating period is presented. Drag
correlation for free-falling particles i n still fluids relates
relative velocity of gas and solids t o force required t o
overcome inertia, gravity, and solids friction.

M e a s u r e m e n t s of total pressure drop and static pressure drop in t h e transport of solid particles through vertical 0.267- and 0.532-inch inside diameter glass tubes by a n
air stream are reported. Closely sized sand (28- t o 35-, 35to48-,48-t060-, and60- to80- meshTyler),and bothground
and spherical silica-alumina cracking catalyst were used.
Solids circulation rates ranged f r o m 2 t o 54 pounds per
second per square foot a t various constant air rates f r o m
0.9 t o 3 pounds per second per square foot, equivalent t o
12 t o 40 feet per second. Direct measurements of t h e dispersed solids density were made; these permitted t h e
average particle velocity, slip velocity, and solids static
pressure drop t o be calculated. T h e t o t a l pressure drop

0.H.HARIU A N D M.C.MOESTAD
UNIVERSITY OF P E N N S Y L V A N I A . P H I L A D E L P H I A , P A .

E of the inherent advantages of a fluidized system is the


ease with which solids can be added to or removed from the
reaction zone by pneumatic transport. A means of estimating
the pressure drop in the ducts carrying the gas-solid mixture is
essential in designing such a system. This becomes particularly
important in designing small equipment such as pilot units where
the risers (as the vertical ducts are callpd in fluid catalytic
cracking units) may be as small as 0.5 inch inside diameter. In
pipes this size, friction loss betiveen the solid particles and the
pipe walls is a n unknoivn and conceivably a large part of the
total pressure drop. At the other extreme, in the 6-foot inaid?
diameter risers found in commercial fluid catalytic cracking units,
the pressure drop is usually assumed to be entirely due t o the
static head of the gas-solid mixture.
During operation of a full scale fluid cat cracker, the riser
pressure drop serves as a guide in estimating catalyst circulation
iates. I n the operation of a pilot unit, where inore exact measurements can be made, it is found that n-hen a constant carriei aii
rate is maintained in the spent catalyqt riser, a linear calibratioii
can be made between riser pressure drop and 8o11ds circulation
rate for a given cracking catalyst. The iiser pressure diop also
is used to hold the catalyst circulation rate a t a desired value
by means of a differential pressure 1 ecorder-conti oller n-hich
actuates the spent catalyst slide valve.
I n systems where all or part of the reaction tabes place in the
riser, that is, when the reactant gas is used as a carrier, it uould
be desirable to estimate the riser solids concentration and velocitv
so as t o provide a measure of the degrce of contact betn-een the
gas and solid. This again is directly related to the static head
component of the total riser pressure drop.
A knowledge of the solids flow rate, gas velocity, and total
pressure drop in a vertical riser offers no clew as to the relative
values of the static and fiiction components of the total pressure
diop. If a section where acceleration of the particles has ceased
is considered, the mass-average solids velocity or the concentration of solids in the riser also must be known before the static
component can be calculated. A t constant solids f l o ~rate (in
weight per unit of time) and carrier gas velocitv, a coarse heavy
material will move a t a slower velocity than a fine low density
powder. Consequently with the heavy material the static head
rvill be higher than in the case of the light material; the friction

loss d l certainly differ due to the difference in velocity of the


two materials
The object of this investigation was to study the effect of the
variables mentioned in the preceding paragraph on the pressure
drop in two sizes of vertical glass tube.

Present address, Atlantic Refining Company, Philadelphia, Pa.

LITERATURE

Little can be found in the literature on the subject of riser


pressure drops in fluidized systems. Daniels (4), in describing
the hydraulics of a fluid catalrtic cracking unit, mentions that
riser gas velocities of 15 to 30 feet per second are requiird and
that pressure drop in the riser.. is estimated commonly by assuining i t equal to the static head requirement and disregarding an>
fluid friction. He calculates the density of the gas-solid mixture
by multiplying the flow density (total flow of solid plus gas in
pounds per hour divided bv total volume in cubic feet per houi)
by an assumed slip factor of 2.0, equivalent to taking the velocitv
of the particles of solid as one half the velocity of the gas. For
example, with a carrier gas velocitv of 20 feet per second, a slip
factor oi 2 would requiie the d i p velocity t o be 10 feet per second.
I n viex of the fact that the average free-falling velocity of
the catalyst particles in the carrier gas is of the order of 1 foot
per second, the slip factor should be about 20 19 or onlv slightly
over 1.0 unless there is corisidei able friction betireen the particles
and the riser lvall.
Dalla Valle (3) presents a summary of most of the literature
pertaining to the theory of particle transport prior to 1940.
The emphasis up to that time n as on the pneumatic handling of
giain He developed equations for the air 7-elocity required to
transport particles. The banie result can be arrived a t by use
of the inore general and fundaniental correlation of drag coefficient against Reynolds number, the data for which are suinmarized by Lapple and Shepherd (8). Cramp ($) presents an
equation, which appears fundamentally sound, for calculating
the total pressure drop in a pneumatic conveyer. Methods
developed by Hudson (5, 6) for the design of pneumatic conveyers are strictly empirical. Jennings (7) shows a method of
calculating the accelerating distance in vertical transport. Chatley (I), in calculating the poiver requirements of a grain convever, recognizes that an additional term should be added for
friction between the solids and the pipe wall, but states that no
information is available for estimating this.

1148

INDUSTRIAL A N D ENGINEERING CHEMISTRY

lune 1949

I n a paper recently published, Vogt and White ( I d ) report


measurements of pressure drop in vertical and horizontal pipe
carrying suspensions of solids in air. They derive a correlation
in terms of the ratio of pressure drop with the suspension flowing
t o t h a t obtained with the carrier gas alone, the weight ratio of
solids flow t o gas flow-, and properties of the solid and gas. Their
analysis is difficult of interpretation for vertical risers because the
allowance for static head is not clearly indicated. No measurements of particle velocity were made. For reasons shown above
the present authors believe it is desirable to have a method of
estimating both solids static head and solids friction pressure
drop. Both these components of pressure drop are taken into
account in the theoretical analysis which follows.
THEORY

UsPds

(1)

Tn the experimental work, uswas calculated from measurements of

G, and

Pds

12/62.3 = 0.1925pd,L,

(2)

01'

Ap,/L

(3)

0.1925 Gg/u8

Expressing this as pressure drop per foot per unit of mass velocity,

Apn/LG, = 0 . 1 9 2 5 / ~ ,

(4)

An assumption is now made that the Fanning friction equation


(IO)will apply to the energy lost by the particles in striking the
riser wall and in impacts between particles. Then the solids
friction loss,
(5)

Ap.f8 = 2 f,Lu3pdS(O.1925)/gD,

Zxpressing pds in terms of us

An analysis of measurements of all the pressure drops encountered in a transport system consisting of a horizontal pipe,
a bend, and a vertical section would be a complex problem.
Particle velocities probably would be different in the horizontal
and vertical sections. The nature of the friction loss between
particles and the pipe wall varies between the horizontal section,
t h e bend, and the vertical pipe. I n the first, the weight of the
particles probably concentrates most of the friction on the lower
half of the pipe; in the second case centrifugal force has a large
effect on the friction; and in the vertical section friction losses are
due t o the particles striking the walls and other particles. Static
head is important in the vertical tube but is zero in the horizontal. Kinetic energy changes are important: initial acceleration a t the point of entry of the solids requires energy which must
be supplied by the carrier gas; at the bend, and for some distance above it, the particles must be accelerated upward, resulting
in an energy transfer from the gas to the solids over and above
t h a t required for solids friction and potential energy change;
if the over-all pressure drop is a n appreciable fraction of the
initial static pressure, acceleration of both the gas and particles,
due to the increase in specific volume of the gas, must be accounted for in the energy balance.
Consider the ideal vertical riser in which uniform spheres are
moving upward at a constant mass-average velocity of us feet
per second, carried by a gas moving a t ug. To support the particles, the gas must slip past them at a relative velocity (slip
velocity) of Au, equal t o (u,- us),and exert a force, F p pounds,
on each particle. The drag correlation, as shown below, relates
Au t o F,. This force is equal to the weight of the particle
whether i t is standing still or moving a t a constant velocity if there
a r e no external forces acting on the particle, such as friction
against the tube walls. Then 4u will be the free-falling velocity
of the spheres. I n the presence of external retarding forces on
the particles, brought about by collisions with the walls and
between particles, a somewhat higher slip velocity is required.
I n the range of slip velocities covered in most of the present experimental work, F , varies approximately as ( A U ) ~ . ~ .
The mass rate of flow of solids per unit cross section of riser
can be designated G, pounds per (square foot) (second). Because
of the slippage, the true dispersed solids density, p d s (in pounds of
solids per cubic foot), is higher than the flow density, which has
been defined above as total weight flowing per hour divided by
total volume of flow per hour. The following rigorous relation
is used in the derivation below:
Ga =

APz =

1149

Pds.

Over L vertical feet of riser, the pressure drop in the carrier


gas due to supporting the weight of the dispersed solids can be
regarded as a solids static head, equal t o L feet of a fluid of density
pds.
Changing t o a unit of pressure commonly used in fluidized
systems, the solids static head in inches of water is

AptJ = 0.1925 X 2f,Lu,2 G,lgD;lc,, or

Ap/,/LG, = 0.1925 X 2 fszcS/gDr

(6)

(7)

A manometer connected across length L will read 4 p , the sum


of a pressure drop due to gas friction against the walls, ( A p ~ u ) ,
plus the solids friction loss plus the solids static head, or
AP

Ap/o

f A ~ j a+

(8)

ATJ~

The static head of the gas phase will not be indicated if the
manometer lines are filled with a gas of the same density. If
the volumetric concentration of particles is low, the true velocity
of the gas will not differ appreciably from the superficial velocity.
The surface area of the riser in contact with the gas is affected
only t o a slight degree by the presence of the particles. Consequently, i t is reasonable t o assume t h a t the gas friction loss is
the same as if the solids were absent.
The total solids pressure drop, from Equation 8, is
Aps

A p - Ap~p/o= AP,

4-

(9)

Apfa

Combining Equations 4 and 7 ,

Apa/LGs = 0.1925 [I/'us 4- (2fa/gDr) X

(%)I

(10)

Equation 10 is one of the basic equations used in the analysis


of the experimental data. It is in accordance with earlier observations t h a t A p / L is a linear function of G,, with its origin
at A p f g / L and G, = 0, if us and fa can be considered constant.
Here Ap,/LG, represents the slope of a plot of A p / L against
G,. The present experimental data verify this linear relation
(Figures 6 and 7 ) . An interesting feature of Equation 10 is t h a t
it is entirely independent of the gas velocity and the gas and
solid properties, except for the possible effect of these variables
on the Fanning friction factor for the solids, fs.
-4second relation between A..o,/LG, and u8is necessary t o solve
for these values at a given set of operating conditions. This is
derived from the drag correlation (8, IZ), which relates the relative velocity of the gas and solids, Au, to the force exerted on the
particles by the gas. By this correlation, the force in pounds on a
particle is

F , = Cp,( A ~ ) ~ . 4 , / 92 =

7rCpo(

Au)'D;/8 g

(11)

where C is a function of the Reynolds number, Dp(A u ) p , / ~ , .


The number of particles in L feet of riser is

The total force on all the particles is

F = ( T C P AAUI2Dg/8 9 ) X (6pdsLArl~D3,pp)

(13)

The total force on the particles divided by the area of the tube
and L equals the pressure drop per foot in pounds/square foot, or

F I L A , = 3 CP&AU)'Pds/4 gDpP,
Converting to inches of water and substituting G8/uIfor p d s ,

(14)

INDUSTRIAL AND ENGINEERING CHEMISTRY

1150
Ip,/L

0.1925 X 3

Cp,(

AZL)~G,/-I
gDnpP1/*or

Ap,;LG, = (0.0045/Dp)( p o / p p ) I ( Au)'/u81 C'

(15)
(ltij

Simult'aneous solution of Equations 10 and I 6 anioiiii ts to a


force-balance in which t,he force required t o support the particles
and overcome solids friction, as specified by 1;quatioii 10, is
balanced by the force exerted by the gas oii the partic:lw as s h o ~ i i
by Equation 16. To relate A u to ul,
t'he gas velocity must be
specified. Then A u = uo - us. For a single smoot,h sphere,
C will vary with Au as showi by the curve oi Figure 10, ivhrrc
C is plotted against Re.
A graphical represent'atiori of Equations 10 aiid 1 6 aids in
explaining their significance. I n Figure 2, lines A C and .2D are
calculated from Equation 10, assuming j equal to 0.004 and
using the two riser diameters used in the clxperimxital ~ v o ~ k .
Line B R is the static head component of ICquatiori 10 arid is
independent of riser diameter. Ap,/LC, approac1it.s infinit?,
as 71r approaches zero, as Ap,/LG, is proportional to 1./1/~. A s
ug increases, ap,/LG, reaches a minimum aiid t,heri bibcomes
greater when the solids friction loss increases iiiore rapidly t hari
the static head decreases. With increasing diamctcr the minimum moves to the right. Line EG is calculated from Ikpatioii
16 for 0.00165 foot dianieter spheres with a density of 165 pounds
per cubic foot (the size and density of one of the sands used in the
expwinient) carried by a 25 foot per second air stream a t atinospheric conditions. The curve for free-falling sphcws in Figure 10
is assumed to hold. Point F is the force balancc point for the
larger riser, determining immediately the values oi'
arid Au,
as well as ap,/LG,, which is the suni of ap,/LG, arid ~ p j , / l , G . .
Thr: ordinate a t point H is Ap,/LG,, t,he static hc,ad component,
wht:reas the solids friction drop per unit inass velocity Apja//,G.>,
is represented by the distance H F . Comparison of thc iritersections G arid P shows the effecl o i decreasing thc size of thv
riser a t a constant air velocity. The solids velocity, u . , tlrcreases slightly in going from F to G, resulting iii a soinr:n-hat
higher static head; the friction drop practically doubles. If
the riser dianieter is increased to the point, that solids friction is
negligible, the line for total pressure drop such as A 1) o r ,IC becomes practically coincident with line A B . The in1
drops to J and the slip velocity, 25 - 11.9 = 13.1feet pw sc:corid,
corresponds to the free-falling velocity of the spheres. This citn
be considered t o be approximately t,rue for large risers.
An expression for the mechanical efficiency of this ideal riser
now can be dcveloped. The powcr output is cyii:iI to thcl P t l T 2
at v-hic'l lift work is done on the solids, or
EIp. output = G,A,L/SSO

(17)

T h o 1m\ver input is t h a t lost by the air aiid cdii be ~ ~ o ~ ) r i ~ s cI)?


iittd

Hp. input

&pug AT/(0.1925) X (550)

(18)

R-liwe Ap = Spja
Apj\
A p z in inches of wat,er. T h e
mechanical efficiency, hp. outlhp. in, is a maximuin when
riser diameter is so largc as t o niake air arid solid friction lo
rrcgligihlr, so tha't A p = A p = pd.L(O.1 923). Thc~l~c~fol*i?
Fcff,",ex

G,,I J,/,; L u g . 1
/,*

j.

(19)

It must be remembered that this analysis is concerned o i i l v


with a vertical section and one in which acceleration of gm rtntl
solids is negligible.
PARTICLE ACCELERATION

Variatioris iri the valurs of fs calculated from the expoi iiiiriii ~ i l


data by Equation 10 indicated the possibility that the p a r t w h ~ ~
had not reached an equilibrium velocity before entering thtl t
section, but were still accelerating. Therefore, the residual solid5
piessure drop, after subtracting the measured static hcad, woulti
be due not only to the foice everted by the gas on the partic I(
to overcome solids iiictiorial forces, but also t o the force ieq
to accelerate the particle.;. In an effort to distinguish bct
the frictional and acceltliatioii components, the following niai 11( matical treatment was derived.
Applied to ono pa1 ticlr, thc i i c ' t u p w x d acceleration ( * A I I i l l
t ~ pcscled
r
as

~ v l i t wthe first term oii the right t h upmard acceleration duc, I ( J


the diag force of the cariicr gas and the second and third ter~ii.
a i r the domnnard accelerations tluv to gravity and the frictioiial
force pLlr particlc, respectiwly. If fs and the empirical rclaLioii
twtweeii ( A u ) and C' arc. kiion 11, 101 a specified gas velocit\
/ i o ,a can b(>calculated for all valuc.5 o r 11, from zero to the poilit
u-licire a = 0. T h e lattcr c . o r r i y ) o n d ~to the equilibrium solitli
vrlocj ty.
T h r time-velocity and dislaiiuc-vt,loc.ity relations caii be 011t a i n d by giaphicnl intcyqi atioii of the L01101\ ing tv,o equatioiis:
i2:3i

I-k~uatiori23 is iritegratcd by plotting I ,'CL (as ordinatt.) agaiii-~


0,is calculated by TGquation 22. T l i e time rcquircd to w : t ( > l i
velocity u2 from ul is thc area under thc curve. Similarly, l:qii:ition 24 is integrated by plot,tiiig u,/a 8s ordinate against u s .
Finally, Equation 10 can bo expressed differentially t o iricliitlr~
t l i ~prttssure drop rryuircid lo :tc:c:i~lc,ratethe particles:
lis.

(25)

Hy t,his equation,

= U9/i/[,

( 20 1

~i,/ic~

cari tjc piottod as a function of

iii.

ii'

(2)

can t,hrii be plotted as a function of distance, 1,. 11 i \

apparent that when the acceleration approaches zero, us ai)proachcs a constant equilibrium value and Equation 25 bwoiii(+
identical with Equation 10.
Equation 25 shows that the p
urc drop per foot can be aik
cutiaf.inely high value a t a point
ere us is small because both

largo.

f i i ~

(k)

arid the acceleration pressure drop

Th(~or(.t,ically,
~l.rcrc,I L ~is zero,

ih(2)

--I

((2

is infinite,

Portunatrly this oarinoh occur actually, sirice for a finit,c G, i u i d


u s , the solids corici~iitr~atioiipds would be iiifinitc:, w h i r ~ l ~

Thus, the iiiaximum mecliariical efficiency i i i 111


accclcrating section of riser is simply the invrrrc. o
which is defined as
Slip factor =

i,"($1)

the varhtion of u with I,&,is kiiovn by Equation 2 2 . Sin(:(>t l i c ,


relat,ion botumm us and /, (:an be derived from Equation 24,

t I1 ti s t at>icIh car1

.k:ff.,,,,,

Vol. 41, No. 6

%<TO

is physically impossible.
(21 1

T h e nicc1iaiiic:iil ene
loss is due to turbulriiw r:tt.at,eti by the
gas slipping past the particles, arid appears as heat. Tlie
efTicic,iicy ~ v i l lbe lowered by air and solids friction losws.

Siiicc the

'-

g21s

term applies to liorimiit til

as \vel1 as vertical ti~aiispoi~i,


this analysis helps in explairiiiig
the high pressure drops and the turbulent conditions normally
t:ncountered a t the poiiit, wherc the solids are introduced into t,hc,
carrier stream. It also indicates that the acceleration p
drop might be decreased hy mcchmically imparting to t'he solitlh

INDUSTRIAL AND ENGINEERING CHEMISTRY

June 1949

a velocity in the direction of the gas stream before the two are
mixed.

This might, for example, be done by a curved nozzle.


APPARATUS

The apparatus shown in Figure 1 was designed to introduce


continuously into a metered air stream an adjustable flow of solid
particles, with facilities for mcasuring the solid circulation rate,
the pressure drop across a length of vertical riser carrying the
gas-solid suspension, and the weight of solids contained at any
instant in a known volume of riser. Dimensions of the apparatus
are given in Table I.

Table I .
4

H.

Dimensions of Apparatus

Storage section
Filter
Cyclone
Inner tube
Outer tube
Storage t u b e
Combined length, D a n d
Lowei tube, inside diam.
Pressure taps
P-1 t o P-2
P-1 t o P-3
Inside diainetei
Standpipe
Length
Inside dianirtcr

Dimension, Inclirs
10 x 10
1 5 (11))
2.5 (1.T))

Riser section

J, I.D.

I to L
1 t o P-4
P-4to P.5
Taps, i d.
Plug valve poits, .I a n d L
Straightening section befoie \al\ e I

AIR

Figure

Apparatus

60
1 82

0 126
16

1.26
Dimensions, Inches
--______
Tube 1
Tube 2
267
5

0
125
25

eter 7 served as a measure of


the solids flow rate,
Two brass Parker three-waiplug valves, I and L, and :t
copper tube by-pass, K (approximately the same inside
diameter as J). were used to
trap a sample of solids in thta
riser for a measurement of pda
The sample was removed for
weighing by: breaking a closefitting rubber sleeve connectioi~
between I and P-4; stopping
the solids flow through B .
connecting the short nipple extending up from Z (containing
the solids) t o a small cyclone
separator fitted with a weighing
flask and a n internal steel wool
filter (not shown in Figure I ) ,
and turning valve I so t h a t the
carrier air forced the sanlplr~
into the flask.
PROCEDURE

6
23

0
41
5
32
0
0
6

1151

0 532

53 9

9
40.1
0.12:
0 5
7

The materials shown in Table I1 were sized with Tyler standard


screens in a Ro-Tap machine, allowing 20 minutes for a small
batch. The sands were further screened by hand until i t appeared t h a t the amount passing the smaller screen was negligiblL
During operation thv
cracking catalysts wen
found to contain a small
amount of fines; thc
ground catalyst contained
more than the microspheroidal, as evidenced
by the amount of dust
rising t o the filter H.
Densities of the sand
particles were measured
by water displacement iii
EI
50-ml. pycnometer.
Since the catalyst particles
are porous, and a measuielo
l5
3G 35
ment of their absolutcl
densities and por0sitit.i
ua, Ft./Sec.
scope
was
considered
of this work,
beyond
v a thv
~u~,
Figure 2. Plot of Equations ,o
and 16
quoted by Webb ( I S ) weie
A6 is static head component of Equawed in calculatillg
t i o n 10: AC a n d AD are plots for t w o
riser diameters, assuming fs = 0.004.
particle density.
The,
EG is plot of Equation 16 for m a t e r i a l
A w i t h gas velocity of 25 ft./sec.
average particle diametei ,

The fluidized bed storage section consists of glass tubes F and D


connected by a valve, E, a conical brass screen with a coneshaped copper seat. The lower end of D is fitted to a brass tube
standpipe, C, at the bottom of which is located slide valve B used
to control the solid circulation rate. The slide valve was constructed from a 0.5-inch gate valve by replacing the gate with a
machined plunger fitting into a cylinder. The flow of solids was
controlled by varying the opening between the end of the plunger
and the apex of a triangular hole in the wall of the cylinder.
A stream of air for fluidizing the bed of solids in D was intro
duced through nozzle F-2. -4second stream entered F-I, at fi
rate adjusted t o maintain a constant positive pressure drop across
the slide valve as indicated by manometer 8. Manometer 6
indicated the pressure drop through a fixed length of the fluidized
bed, between taps P-1 and P-2, and served as a guide in setting
the air rate through F-2 Starting with a n unfluidized bed, as
F-1 and F-2 were increased, manometer 6 showed a n increasing
reading until at the fluidization point it reached a maximum and
indicated the truebed density. At higher fluidizing air rates the
density reading decreased due to expansion of the bed.
Manometer 7 measured the pressure between P-1
and a point above the bed, P-3. It is in effect a
level meter since it measures the weight of solids
above point P-I. Small bleed streams of air were
Table I I . Solids Properties
added through bubblers t o pressure t a s P-1, P-2,
.
I
a n d B = O t t a \ % asand
and P-3 to prevent their blockage by soEd particles.
C a n d D = Sea sand
Two calibrated orifices were used at A to cover
E = Micros heroidal oraching. catalyst
the range of carrier velocities desired in the two sizes
F = Grouiicr cracking ratalyst
of glass riser J. During a run, the carrier air picked
h ~ ~ ~ ~ ~ i ~ l A
B
C
D
h
I
up the solid particles falling through the slide valve
Tyler scregll sise, mesh
and transported them through riser J i n t o cyclone G.
1007 thru
35
48
BO
100
LOO
The solids fell through tube F and valve E (open)
1004o n
28
35
48
BO
80
200
200
D ~ arithmetic
,
av. diameter, 0.00165 0.00117 o OOOYO o 00070 o O O O ~ B o ooo41
back into the fluidized bed in D whereas the carrier
feet
air plus blped air and fluidizing air left the system
density, 1b leu
ft.
165
165
165
169
615
61
through the heavy cloth filter on the top of cone H .
Manometers 4 and 5 measured the static pressure
Ki, for airb
808
574
141
343
176
170
and pressure drop in the riser 3.
The solid circulation rate was measured by
K'* for 'Irc
102
726
658
434
2 2
2 2
closing valve E which caused a fluidized bed of
ut, calculated free-fall velocity 12 8
9 2
7 2
5 2
1 0
I 0
particles to build up in tube F. The rate remained
in
ft*'sec.d
Calculated, assuming absolute density = 2 33 g./cc a n d poie 5olunie = 0.6 cc
unchanged because the total weight of solids above
b Used in calculating drag; Ki = 222 D p p p / p g ; P O = 0.075 Ib /cu f t .
C K2 = D p p g / p g so Re = K2( Au), PO = 0.075,PO = 0 018 X 0 000672.
the slide valve remained constant, ensuring a
constant slide valve A p . As the level dropped
For spherical particles; for rough particles, ut is lower.
in D , the rate of change in the reading of manom-

*'

INDUSTRIAL AND ENGINEERING CHEMISTRY

1152

Vol. 41, No. 6

D,, was taken as


the arithmeticmean
of the tmo screen

openings.
Photomicrographs of the
six materials are
shown in Figure 13.
Each scale division
I
corresponds to 100
microns.
I T ith each mate40
0
I
2
3
4
rial, a t one setting
Time, Min.
of the orifice meter
Figure 3. Measurement of Flow
2-that is, constant
Rate of Solids
carrier air ratea series of runs
was made coveiing"as wide a range of solids circulation rates as possible. Two
i o four carrier air rates were used with each material in each ot
t h e two risers. Tables 111 and I T show the range of variables
covered
The procedure employed in a run n a s as follows: after the
material was charged to the unit, F-2 vias adjusted to fluidize
The bed in D without excessive slugging. The carrier rate was
adjusted to the desired pressure drop on meter 2 and slide valve
~~

4p Across Valve, i n . H g

Figure 4.

Gas Leakage through Slide Valve

B n a y opened t o start the circulation of solids. Valves I and


L n-ere in the position shown in Figure 1, 60 the gas-solid mixture
flowed through riser J . The air rate to F-1 was adjusted t o
maintain a desired constant pressure drop across the slide valve
as shown by manometer 8. The air rate to F-2 was finally adjusted t o as near the minimum fluidization point as possible t o
reduce slugging and fluctuations in the level meter 7. Readings
n-ere taken of manometers 2 through 8.

~~

Table I I I.

&?/

Smoothed Data and Calculated Results for 0.267-Inch I.D. Riser


G.,

Ap,

)LS<

M S ,

Grams Cu. Ft.


Solid See. In. HzO Lb.,"r
1.10
1.83
15.7
2.0
A
21.9
2.93
1.76
29.4
21.9
3.0
2.41
4.02
43.0
21.9
4.0
5.08
3.05
5.0
56.5
21.9
3.71
6.18
70.2
6.0
21.9
..
..
..
A v . 21.9
...
0.76
1.27
19.5
3.0
35.3
1.04
1.73
31.7
4.0
35.4
2.18
44.0
1.31
35.5
5.0
1.58
2.63
56.1
35.6
6.0
..
..
..
...
AV. 3 5 . 5
0.42
0.70
13.0
3.0
40.9
1.13
0.68
24.5
41.1
4.0
1.57
0.94
36.0
41.3
5.0
1.20
2.00
47.5
6 .O
41.5
..
..
sv. 4 1 . 2
.j.
1.30
0.78
18.9
2.0
B
21.7
1.26
2.10
32.9
21.7
3.0
2.92
1.75
47.0
21.7
4.0
3.72
2.23
5.0
01.0
21.7
2.71
4.52
6.0
(4.5
21.7
..
..
Av. 2 1 . 7
...
0.62
0.37
10.5
30.7
2.0
1.08
0.65
3.0
22.6
30.7
1.57
0.94
34.6
30.7
4.0
2.03
1.22
46.8
30.7
5.0
1.50
2.50
58.8
6.0
30.7
d v . 30.7
0.45
12.4
0.27
41.2
3.0
0.45
0.75
22.2
4.0
41.3
1.03
0
.
6
2
3
2
.
1
B.0
41.5
1.33
0.80
42.0
41.6
6.0
..
..
..
Av. 4 1 . 4
...
1.40
2.31
20.2
2.0
C
13.3
3.64
31.4
2.18
3.0
13.3
2.96
4.94
43.0
4.0
13.3
3,73
6.21
54.4
13.3
B.0
..
..
Av. 1 3 . 3
..i
1.37
0.82
18.3
20.3
2.0
2.2,s
1.35
32.8
20.3
3.0
3.14
1.88
47.2
20.3
4.0
2.41
61.5
4.02
20.3
5.0
4.90
2.94
20.3
6.0
75.8
..
.&v. 2 0 . 3
...
0.52
29.1
11.1
0.31
2.0
0 . a
0.98
22.9
3.0
29.1
1.45
34.8
0.87
29.1
4.0
1.90
46.7
1.14
29.1
5.0
1.41
2.35
58.5
29.1
6.0
..
Av. 2 9 . 1
...
a Includes pressure drop due t o acceleration of

..

..

..

..

..

..

Lb.1

(Sq.Ft.)
(See.)
11.3
21.1
30.9
40.5
50.4

...

14.0
22.8
31.6
40.3

..

9.3
17.6
25.8
34.1

..

13.5
23.6
33.7
43.7
53.3

7.5
16.2
24.8
33.5
42.1

8.9
15.9
23.0
30.7

..
14.5
22.6
30.8
39.0
13.1
23.5
33.8
44.1
54.3
8.0
16.4
24.9
33.4
42.0

..

u8,

Ft.,'Sec.
6 2
7.2
7.7
8.0
8.2
7.5
11 0
13.2
14.5
15 3
135
13.3
15.6
16.4
17.1
l5,6
10.4
11.2
11 5
11.7
11.8
11.3
12 1
15.0
15.8
16.5
16.8
16.0
20
21.2
22.3
22.6
21.5
6.2
6.2
6.2
6.3
6.2
9.6
10.4
10.8
11.0
11.1
10.6
l5,3
16.7
17.1
17.6
17.8
16.9

the particles.

Apparent
~ p ~ .lPJ,/ua
,
.lPaB
APfa
In. H20 In. HzO In. H2O in. H;O
0.94
0.54
1.46
0.52
1.51
..
2.46
0.95
2.06
..
3.46
1.40
2.61
4.46
1.85
3.17
..
5.46
2.29
.

0.65
0.89
1.12
1.35
I

0.36
0.68
0.81
1.03

..

..

1.37

1.63
2.63
3.63
4.63

0.98
1.74
2.51
3.28

1.82,

1.18
2.18
3.18
4.18

0.82
1.60
2.37
3.15

..
..
..
..

..
..
..
..

..

0.67
1.08
1.50
1.91
2.32

0.54

..

1.46
2.46
3.46
4.46
5.46

..

..

1.OG

0.94

0.w

..

2.94
3.94
4.94

2.13
2.90
3.66

32
0.56
0.81
1.04
1.28

0.23
0.39
0.53
0.68

..

..
..

..

..

..

..

..

1.94

..

0.79
1.38
1.06
2.85
3.14

i.is

Frictiona
AU,

fs
Ft./Sec.
0,0052
15.7
0,0044
14.7
0.0041
14.3
0.0040
13.9
13.7
0.0039
,, ,
14.4
0.0043
24.3
0,0040
22.2
0,0038
21.0
0.0037
20.3
22.0
27.6
0,0046
0,0041 25.5
0.0039
24.9
0.0037
24.4
, . ,
25.6
0.0039
11.3
0,0038
10.5
0,0036
10.2
0.0035
10.0
0,0035
9.9
...
10.4
0.0047
o.0040 1 5 . 7
0,0038
14.9
0,0037
14.2
0,0036
13.9
14.7
0.0037
21.2
0,0037
20.1
0,0036
19.3
0,0036
19.0
., ,
19.9
0,0046
7.1
7.1
0,0046
0,0046
7.1
7.0
0.0045

1.17
2.17
3.17
4.17

0 94
1.78
2.64
3.49

..

..

1.80
2.80
3.80
4.80

0.60
0.93
1.26
1.60

..

..

..

....

0.71
1.16
1.61
2.06
2.52

0.46

..
.. ..
..
..

1.54

2.54
3.54
4.54
5.54

0.83
1.38
1.93
2.48
3.02

0.0048
0,0039
0,0037
0,0036
0.0033

0.27
0.50
0.76
0.98
1.21

0.94

1.06
2.06
3.06

, .

1.20
1.37
2.54
3.20

..

1.83

..
..

...

..

0.20

..

..
. I

..

4.06

5.06

0.79
1.56
2.31
3.08
3.85

...

0.0046
0.0040
0.0038
0,0037
0.0036

U~,'V.~

3.54
3.05
2.85
2.75
2.69

Re

.. .
., ,
.. .
, . .
...

__

1.27

3.21
2.68
2.45
2.33

...
. ..
. .,
, ..

..
.. . ,.
..
. .
,

147

...
.,,
,.,
...

..

0.98

3 07
2.64
2.52
2.42

..,
...

..

0.89

2.09
1.94
1.89
1.86
1.84

. .,
..,
...
.,.
...

...
..,
...
..,
, , .

..

2.34
..

74

2.04

1.94
1.86
1.83
, ,

2.08
1.94
1.86
1.84

..,
...

, ,

.
. ..

.,,

1.91

..,
.,,
..,
.,,

224

...
...

.. .. ..
262

. ..

..

, ,

...
, ..
107

...
,,,

...

...

..

1.57

1.87
1.87
1.87
1.90

., ,
..,
...

..

..

...

., ,
...
. ..
, . .

2.54

40

10.7
9.9
9.5
9.3
9.2
9.7
13.8
12.4
12.0

2.12
1.95
1.88
1.84
1.83

...
., .

..,
...
.,.

...
...
, . ,
...
. ..

2.00

54

11.5

11.3
12.2

..

1.90
1.74
1.70
1.65
1.63

..

..

145

.
.
.
.

.
.

.
2.32

68

-\pdLGa
0 0487
0.0439

0.0420
0.0413
0.0406
0.0435
0.0436
0,0432
0.0430
0,0430
0,0432
0.0475
0.0464
0.0462
0.0460
0.0466
0.0405
0.0390
0.0384
0,0383
0,0383
0.0389
0.0470
0.0448
0.0445
0.0441
0,0440
0.0449
0.0492
0.0511
0.0517
0.0519
0.0508
0.0405
0.0464
0.0462
0.0461
0 0463
0.0440
0.0405
0.0392
0,0385
0.0382
0.0401
0.04Y6
0.0470
0.0460
0,0456
0.0461
0,0467

(Concluded on p a g e 1163)

INDUSTRIAL AND ENGINEERING CHEMISTRY

June 1949

The solids circulation rate, ws,


was then measured by closing
valve E and reading manometer 7 a t precisely 15-second or 30second intervals. These level readings were plotted during the
run against time, as shown in Figure 3. The best line drawn
through the points gave the average rate of change of manometer 7 in centimeters per minute. Manometer 7 was previously calibrated for each of the materials by weighing the solids
removed through the slide valve and noting the change in the
meter reading. The calibrations checked the calculated values
within 1%. Figure 3 shows three circulation rate determinations.
The slope of each line times the calibration factor gives the
solids rate in pounds per hour.
After the solids circulation rate was determined] valve E was
opened, allowing the solids to flow back into D. With no changes
made in the settings, valves I and L were turned simultaneously
to by-pass the flow through K and isolate a sample of solids in
J . After removing the sample as described above, it was weighed
to the nearest 0.01 gram.
It was found t h a t there was an appreciable leakage of air
down through the slide valve, adding an unknown amount to
the measured carrier gas. A method was devised to calibrate
the leak; a carbon dioxide stream was substituted for the air a t
F-1, the aeration point above the slide valve. Orsat analyses
of the riser gas while solids were flowing through the valve gave a
measure of the amount of leakage as a function of slide valve
A p . These measurements are shown in Figure 4. With all four
sands (A through D) the leakage was found to depend only on

1153

---- CALCULATED
-OBSERVEJ

the slide valve, A p ,

opening. This is because at the aeration rates used, the


standpipe remained
unfluidized and the
main resistance to
downward flow of
aeration gas from
50
100 200
400
point F-1 was in the
Standard Cu. Ft./Hr. (32')
2-inch depth of
particles. With the
Figure 5. Pressure Drop in
finer and less dense
Absence of Solids
cracking catalysts
no consistent calibration was found. A slight flow of aeration gas in F-1 would
fluidize the standpipe and then the leakage, as measured by the
carbon dioxide method, appeared to vary with valve pressure
drop, the valve opening, and the fluidized density of the material
in the standpipe. At a constant valve pressure drop the leakage
was approximately proportional to the solids circulation rate.
Use of this correction, as shown in Tables I11 and IV, introduced
a n uncertainty which could not be evaluated.
The friction drop due t o air flow alone was measured a t various
air rates for both glass tubes used as risers; these data are presented in Figure 5 . Some of the points were taken over a period

Smoothed Data and Calculated Results for 0.267-Inch I.D. Riser (Concluded)
Gs, Lb./
Apparent Frictiona
Av.Is,Ct.7 ( S S . Ft.) us
APZ,
Anfa
APS
Apjs
Au ,
Gram:: Cu. Ft
(Sec.) Ft./S'ec. In. Hz0 I n . H i 0 I n . Hi0 in. HzO
f.
Ft./Sec. ua/uz
C
0.26
0.43
0.22
9.3
21.6
1.32
1.68
1.10
0.0038
18.3
1.85
...
0.44
0.73
15.1
20.8
..
2.32
0.38
1.94
0.0043
19.1
1.92
...

Table I l l .
Solid

Fut/
Sec.

AP,
we,
In. Hz0 Lb./Hr.

39.9
39.9
40.0
40.1
Av. 4 0 . 0

3.0
4.0
6.0
6.0

12.2
12.2
12.2
12.2
Av. 12.2

2.0
3.0
4.0
5.0

19.7
19.7
19.7
19.7
Av. 19.7

2.0
3.0
4.0
5.0

28.7
28.7
28.7
28.7
A v . 28.7

2.0
3.0
4.0

..

..
Av. 1 9 . 3 '

...

...

...

5.0

...

1.0
1.5
2.0
2.25

...

1.25
1.5
2.0
2.25

.4v. 27, 8

..
..
..

Ax-. 19.6C

45..

28.0C

...
1.o

1.5
2.0
2.5

12.9
21.2
29.5
37.8

..

20.2
33.2
46.3
59.4

..

18.5
31.2
44.0
57.0

..

11 6
22.3
33.1
43.9

..

4.6
10.5
16.7
15.8

..

2.6
4.5
8.3
10.2

..

6.3
12,8
19.4
26.0

...

..

1.25
1.5
2.0
2.25

4.0
6.8
12,6
15.3

...

..

0.63
0.81

..

1.38
2.17
2.96
3.75

..

0.77
1.26
1.76
2.25

..

0.32
0.60
0.88
1.16

..

0.15
0.31
0.48
0.56

..

0.03
0.06
0.14
0.18

..

1.05
1.35

..

2.30
3.62
4.94
6.25

..

1.28
2.10
2.93
3.75

..

0.53
1.00
1.47
1.93

..

0.25
0.52
0.80
0.93

..

0.05
0.10
0.23
0.30

..

21.2
27.1

..

14.5
23.8
33.1
42.5

..

13.2
22.3
31.5
40.8
. I

8.3
16.0
23.7
31.4

..

3.3
7.5
12.0
14.2

.*

1.9
3.2
5.9
7.3

..

0.16
0.32
0.49
0.65

0.27
0.53
0.82
1.08

..

..

..

0.06
0.10
0.18
0.22

0.10
0.17
0.30
0.37

2.9
4 9
9.0
11.o

..

..

4.5
9.2
13.9
18.6

..

20.2
20.1
20.7

0.54
0.65

6.3
6.6
6.7
6.8
6.6
10.3
10.6
10.8
10.9
10.7

0.66
1.os
1.51
1.93

..

..

3.32
4.32

2.78
3.63

..

..

1.18
1.86
2.54
3.21

0.19

..

1.81
2.81
3.81
4.81

..

..

..

..

1.56

0.90
1.48
2.05
2.63

..

..

..
0.44

..

..

0.63
0.95
1.27
1.60

2 56

..

.. ..

..

..

..

15.6
16.0
16.1
16.3
16.0

0.27
0.51
0.76
0.99

0.92

0.81
1.57
2.32
3.09

, .

..

1.08
2.08
3.08
4.08

13
14.4
15.0
15.3
14.9

0.13
0.27
0.41
0.48

0.38
0.40
0.42
0.44

..

0.03
0.05
0.12
0.16

28'
24
25

..

..

..

.. ..

t .

0.82
0.83
0.85
0.86

..

16.7
17.3
17.0
17.2
17.1

0.14
0.27
0.42
0.56

..

..

29
29
30
30
29 5

0.05
0.09
0.15
0.19

0.82
0.83
0.86
0.86

..

0.38
0.40
0.42
0.44

..

3.56
4.56

..

0.62
1.10
1.58
1.81

..

0.43
0.67
1.15
1.39

..

0.62
1.10
1.58
2.06

..

0.43
0.67
1.15
1.39
. I

..

0.49
0.83
1.17
1.33
0.40
0.62
1.03
1.23

..

0.48
0.83
1.16
1.50

..

0.38
0.58
1.00
1.20

..

0.0045
0.0046

....

0,0048
0,0042
0.0040
0,0038

.. .

, .

1.40

5.5

1.53
1.85
1.82
1.75

...
...

...

3.2

23

...

...

f.2
a.a

5.4
5.6

..

1.51
1.88
1.82
1.81

0,0043
0.0043
0,0042
0.0042

13.1
12.7
12.6
12.4
12.7

. . ..

0.0052
0,0045
0,0044
0.0068
0.0054
0,0048
0,0047

....

0,0044
0.0037
0.0034
0,0032

....

0.0038
0.0035
0,0032
0,0031

...

..,

1.98
1.99

9.4
9.1
8.5
8.8
9.0

t . .

...

...

15.8
20.0
19.3

0.0046
0.0044
0,0042
0.0041

.. .

Re

...

..

...

..

1.95

35

...

..
b

..

..

..
..
..
..

..

..
..

..
..
.*

2.5

1.15

...

...

55

...

...

...

...

...
...

9.7

..,

...

, . .

...

...

...

...

..,

...

..

, . .

...

...

...

...

..,

20

5.5

..

..

..

.,.

1.65

..

..

...

...

..

. I .

. . ,

...
...

..

...

1.84
1.75
1.78
1.76

7.0

..

...

...

1.3

. ,

...

,.

..

...

...

4.4

.. ..

, . .

..,

...
I..

...
..,

...

...
.

A
17.8
2.5
23.5
1.77
2.95
16.8
5.7
1.52
0.35
2.15
0.63
0.0046
12.1
3.12
...
3.0
30.5
2.20
3.67
21.9
(GO, 17.8
6 0
1.88
..
2.65
0.77
0,0041
11.8
2.98
...
45.0
3.03
5.05
car17.8
4.0
32.2
6.4
2.60
3.66
..
1.05
0,0036
11.4
2.80
...
4.75
55.8
3.67
6.11
rier) 17.8
40.0
6.5
3.14
4.40
1.26
0.0034
11.3
2.72
...
Av. 17.8
...
..
..
6.1
..
..
..
...
11.7
1.05,
b Not calculated due t o t h e l o a accuracy of us a n d u~).
6 Estimated because slide-valve leakage could n o t be measured accurately with materials E a n d F ; leakage estimatpd a t 1 cu. f t . / l O lb. wg.

..

..

..

108

...
...

..
..

,.

, . .

...
...
...

...

21.6

Aps/LGs
0.0531
0 . C576
0.0587
0.0897
0,0573
0 , 0468
0.0441
0.0431
0,0424
0.0441
0,0443
0.0430
0.0424
0.0418
0.0429
0.0487
0.0481
0.0481
0,0481
0.048f

....

0.055
0.049
0.048.

0,051
0.088.
0,078.
0.065.
0.063
0.073:

....
0.045.
0.043
0.042
0.043
0.056.
0.051
0.048
0.047
0.051
0.04790.0453
0.0452
0.0412
0.0442

INDUSTRIAL AND ENGINEERING CHEMISTRY

1154

Table IV.

17.1 0.80

17.1
17.0
17.0
17.1
29.5
29.5
29.4
29.3
29.4
40.5
40.4
40.3
40.4
16.6
16.5
16.5
16.4
10.8
40.0
39.9
39.8
39.9
16.1
16.1
16.1
16.0
16.1
28.6
28.5
28.4
28.5
15.8
15.8
15.7
1.5.7
1.5.8

28.2
28.2
28.1
28.2
d

14.2

e
1

..

..

0 . 8 0 14.0
1.10 31.6
1.40 49.0
1.70 66.8

..

..

1.40 24.0
1.70 39.5
2.00 55.0

..

0.50 1 7 . z
0 . 8 0 35 r?
1 10 5 4 . 0
1.40 72.5

..

..

1.10
7 6
L 4 0 22.0
1.70 36.6

..

.,

0.50 2 0 . 2
0.80 38.0
1.10 5 6 . 4
1 4 0 74.6

2.56
?.SO

0.05
6.30

Smoothed Data a n d Calculated Results for 0.532-Inch I.D. Riser

0.82
1.22
1 69
2.02

3.41
3.48
7.55
9.68

,.

0.73
1.38
2 40
3.20

0,234

0.50
0.77
1.03

2.51
5.66
8.80
11.90

..
0,79
1.32
1.8:;

1.08
2.18
3.28
4 .?8

0.253
0.42
0 5<J
034fi

0.70
1.05
i.~ii

4.31
7.10

'1.87
:j.OY

6.37
9.70
M.o:!

, .

0.24
0.60

0.95
..
1 00
2.0;

3.11
4.17

U.077
0.192
0.303
0.3Y
(].titi
1.00
1 R1

1 3 :
3.95

6 J7
3.63
6 83
10.11
13.29

, .

0.80 16.2
1.10 3 5 . 6
1 . 4 0 53,O

..

0 50 2 2 . 0
0 . 5 0 40.;;
1 10 5 9 . 4
1.40 78.0

0.49
1.07

1.6.5
1,O.j
2.10
L1.j

4 20

0 Id7
0 34c3
0 .iXO

0.237
0.674
1.01
I 3.5

2.91
6.38

9.88
3.93
7 27
10.67
14.00

1 .

0.80 18.0
1 . 1 0 37.0
1.40 06.0

..

0.40
0.60
0.80

..

16.0
32.5
48.8

..

9.3
0.40
0.60 21.0
0.80 33.0

0.49

1.01

1.62

0.62
1.40
2.28

0.1.57
0,324
0.49
O,l9!)
0.450
0.716

3.23
6.65
10.07

2.88
2.84
8.75

4.2

4.5
4.7
4.8
4.6
10.7
11.3
11.4
11.6
11.3
17.0
16.7
16.6
16.7
8.9
9.1

~4.3

9.1
17.:
20.6
21.5
21.lt
11.0
10.4

..

0.23
0.67
0.91

0.050
0.183

0 292

1.67
3.77
5.92

0.53
0.79
1. 0 4
1.30

..
0.13
0.32
0.so
0.66
0.16
0.27
0.38

..

0.22
0.43
0.65
0.91

..
0.03
0.12
0.20
.,

0.23

..

..

..

..
0.

..

..

..

..

0.96
,.

..
..
0.21

..

..

..
0.95

0.57
0.87
1,17
1.47

..

r) 2.1
0 . 34
n 84

1.14

0.44
0.74
1.04

0.29
0.89
0.89

1 ,19

0.l.j
0 43
0,75

..

..

0.20

9.9

..

0.30
n.60
0.90
1.20

10 1

..

..

..

0.10

0.23

(1 27
I). 57
0.87

10.1

18.5
18.6
18.6
18.6
11.7
10.8
10.6
10.1
10.8
20.6
20.6

20.6
' 20.6

0.22
0.34

,.

..

, ,

..

0.13

0.17
0.09
0.22
0.34
0.48

0.28
0.47
0.66

..

..

..

0.04
0.08

..

0.21
0.43
0.65
0.56

0 07.
0.14
0.21
0.28
0.10

0.33
0.55

0.22
0.43
0,6,5
0.87

..

0.10
0.21
0.31

..

0 . 80
0,80
0.90
1 .20

..

..

0,33

0.27
0.57
0 87

0.20

..
..

, .

,.

..

14.5

13.0
12.2

0.13
0.29
0.46

0.16

12.6b

, .

0.05
0.12
0.19

..

..
..
0.29

..

..

..

..

0.05
0.13
0.20

0.29
I .

..

0.24
0.44
0.B4

0.11
0.31
0.81
0 11

0.31
0 .5l

0.0031
0,0036
0,0041
0.0041

12.9
12.6

12 3
12.2
12 3
0.0037
18.8
0.0038
18 2
0.0038 1 8 . 0
0,0039 1 7 . 7
18.2
0,0042 2 3 . 2
0.0044
23.7
0,0045 2 3 . 7
23.7
0,0028
7.7
0.0027
7.4
0.0020
7.3
0.0025 7 . 1
7.1
0 , 0 0 4 7 29.3
0.0045 I%:+
0.0043 1 8 . 3

U"/W*

cy

Re

Lips 1L 0,

4.11
3.81
3.63
3.53

. .

. .

..
2.75
2.61

0.09
0.17
0.23
0.34

, .

2.38
2.42
2.12

0.17
0.33
0.33

1.87
1.82
1,70
1.77

..

2.28

1.94

1.8,5

0.08
0.17
0.25
0.33

5.1
5 7

1.f7
1 .<55

ti.0

0.002R

6.1

1.02

n.0033
0.0033
0.0032

....

0.002
0.0024

0.0024
0,0025

io. L
Q 9
9.8
9.9
4.1
5.0
.5.1

3.3

18fi

..
..
...

. , ,

..,

0.0312

...
...

..,

128

iio

1 . .X
1 .3
1 .5:3

I . :ii

I , 48
I .18
l,,jl

5 0

..

0 06
0 19
0.32

0.0018
0.0027
0,0030

.,

..

...,

..

0 06
0 18
0 31

0 002
0.0021
0 0025

...
..

..

..

...
...

...

...

...

3.76

22

...

. .
, ,

. o

...
20

...
...
..

...
...

0.0281
0.0277
0.027,i
0,0274

0.0277
0.0332
0.0341
0.0342

0.026G
0.0270
0.0264
0.0288
0.0268
0.026.4
0.0273
0,02-L2
0.025:i
0.0258
0.0261
0,0254
0,0306

...

0,0284

3:i

0.0289

..

, .

..

..

. , ,

,j5

...

0.0314
0,0311

0.0338
0.0247
0.0262

2 28

, . .

, .

..

. .

, . .

, .

1 ,1:3

..

32

3.52

. .

...

..

l.(j

..

24

3.J6

7 0

....

..

..

1 . I(<

...
0.0029
0,0022
0.0019

...

...

..

..

242

..,

...

0.18

...

, .

1.38
1, 36

..

...

2,fi3

7 f>
7 7
7.5

1.37

, . ,

. .
...

0 0028
0 0028
0.0030

0.15

..
..

...

0 17
0.36
0 56

0.11

, . .

II

0.75

0.0025
O.O02(j
0,0027

-..

0.79

0.0500
0.0476
0,0462
0.0435
0.0473
0,0287
0.0286
0,0286
0.0286
0,0286
0.0307

..

, . .

2.58
, . ,
2 , .i:i . . .

..

, .

, .

..

21
20.6
20.3
20.2
, .
20.6
0.30
0 OQ6
0.40 11.8
2.12
22
0.61
0 195
0 . 6 0 23.7
1.25 2 1 . 8
d
0.80 35.5
0.92
0.29:
6.37 21.6
20.2
..
..
21.8
lncludes pressure drop due t o acceleratiori of the particles.
Using average of the last t w o nuinhers.
N o t calculated due t o low accuracy of u a a n d up.
Estimatod. by assuining R -lide-val\.i: leakage of 1 CII. f t . / t O lb.
d

19.0
1.10 30.5
1.40 42.2
1.70 53.9

Vol. 41, No. 6

0.0277

...
, . .
...

0.025
0.023
0.022
:3 . : 0.0233
...
0.020
..,
0.025
, . .
0,026
0.023

...
...

...

0.016
0,022
0.024

0.021

?cy

of a month, during Tvhich time a iiuniber of run6 were made with

m i d . The constancy o the data indicates no change in thc


ioughness of the tube walls. Figure 5 also shows the pressure
drops calculated by the Fanning equation using the f line for
smooth tubing ( I O ) . The vertical line on the curve for tube 1
is a t a Reynolds number of 2100, a t the transition point between
\iscous and turbulent flow. The range of air flows used in tubc
2 as well into the turbulent flow region.
o n e series of runs also was made in tubc 1 with material
using rarbon dioxide as a carrier gas.
Arialysk oi the trends of the apparent solids friction factor
j h indicated the possibility that particle acceleration was contributing to the component of pressure drop originally believed
to he entirely solids friction. .\ pressure tap therefore, was initalled on the larger glass riser midway between the original. tapr
fo pcrmit Ap measurements across the individual halves. Measurements were made a t one air velocity with sand A and two
velocities with sand B over the iiiaxiniuin range of solids f l o ~
iatcs. The air velocities were the same as those used in the
original measurements. The ratio of Aps in the lower half to
that in the upper half remained constant for all values of G,
a t R given carrier velocity Data are summarized in Table V.
A\,

DATA A N D C A L C U L A T I O N OF RESULTS

.ill oi the measurements 01 ut6,solids flow rate in pounds p ( > ~


hour, and M S ,the weight in grams of solids trapped between the
two by-pass valves in the riser, wcw plottcd againit ~ p the.
,
observed riser pressure drop for each series of runs. -\ wriw
consisted of measurements a t a constant carrier rate ~ i t hone
inatciial axid riser. Examples of these plots are Figures 6, 7 , 8,
and 9. In no case mas there a deviation trom a straight-line
ivlatiou. I n almost every plot the straight lineis extrapoltttd t o
the eupcrimental value of air friction drop with no solids R o n
ta1rt.n from Figure 5 . The value used for the air rate was thv \urn
of tiicl inctcrcd carrier air and the slide valve leakage. Thr ail
ltiakagr was ralculated by multiplying the observed value for
carbon clioxide a t the experimental slide valve A p , by 1.24The curvc for uo = 35.4 feet per wcond i n
that is, 1/4%%
Figure 8 shows the greatest deviation from straight-line eutrapolation to A p f o ~ It is included to illustrate one possible sourcc
of wror in the data. I n thi3 series, the deviation is believed
due to a turbulence that was observed directly above the 1owc.r
bypass valve I, but below the lower pressure tap, causing a highvr
concentration of solids at that point than elseyhere. Thvit'

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

June 1949

Table V.

29.4

____
~p~ (lower) a
~

determine usat the two values of np,/LG, in the lowrr and upper
halves. The contribution of acceleration to the apparent j 8
2fsua
shown in Table IV was approximated by equating __ t o

Measurements of Pressure Drop in Lower and


Upper Halves of 0.5-Inch Riser

hIateria1

L)

15 8

a,from Equation 25.

28.2

1.47

1 73

2.00

1.19

1 27

1.34

0.81

0.73

0.66

0.0286
0.0340
0,0232

0,0254
0.0322
0.0185

0.0289
0.0387

11 3
10.3
12.5
0 0038
0.0026
0.0012

10 8
10.1
11
0 0024
6

20.6
19.3
22 oo28
1

0.0018
0 0006

0,0018

A P ~ bL
V.1
A P S / L(upper)

A P ~ (av.)
L
A ~ J L G s (av.)b
Lower half
Upper half
ug (av.)a

Lower
UpperC

is apparentb
.r8 due to acceleration
Friction, fa
a Measured.
b From Table
c

QD,

The term a/u, equals du/dL, or appioxi-

Inately (u, upper - us lower)/L, where IJ is one half the length


of the test section. This treatment is not completely rigorous
because i t assumes the average velocity in each half occurs at t h r
mid-point, but it serves the purpose of illustrating the large effect
on pressure drop of what would appear to be a small amount of
acceleration.

(upper)
S

i p S / L (lower)

1155

0.0191

DISCUSSION

The data in Tables I11 and I V are smoothed values from lineai
plots of w,, solids flow rate, and M,, the weight of particles trapped
in a known volume of riser, against Ap, observed riser pressure
drop. A total of 224 measurements of wsand 256 of M,was used
i n the original plots.
In Tables I11 and IV the following relations are shown:

Oo10

IV.

Calriilated by Equation 16.

At a constant carrier gas velocity, ug, the solids velocity, ti,,


and slip velocity Au, can be considered practically constant a t all
values of wa.
With one material in one riser, increasing uo results in a lower
and static pressure drop, A p z , and
Iispersed solids density,
higher apparent solids friction drop, Apf,,, for the same value of
G,. The slip velocity, Au, increases a t the same time.
For one riser, the major effect of decreasing the particle sizc
and density is to decrease the Sl1P velocltY.
As the components of A p B other than static head brconie
qiI1nller, the slip velocity decreases and approaches the free-falling
velocity of the particles. The slip velocities in the larger tube at

ns a virud evidence of wine tuibulence a t this point in all the


lulls, klut in thls partlcular st111es conditions nere evidently such
515 t o inalce the error a maximum.
The slopes of the lines in Eigurc 7 show that at the lowest gas
yclocit), 17.1 feet per second, the, value of I p , / L G , (proportlonal to the inverse of the slope on this Plot) is the highest of the
three vrlocities used. It is less a t ug = 29.5 feet per second and
slightly higher again a t un = 40 4 feet, per second
This is in accordancr n i t h thc theoretical lines AF'C iii
30
Flgulc 2
Itisel static pressure was a linear function of A p
70
also. rill of the original data were smoothed in this
60
manner, so the data arid ealculations shown in Tables I11
and IV are based on interpolation a t even values of
i 50
-IQ.
Values of G, and ,wswere calculated directly
40
from tos and M,, and u j from the equation us =
-I
Gl/pdJ.
Both Apjg and uB are based on the sum of the
g 30
iwtered carrier gas and the slide valve leakage. Equa20
tions 2 and 9 were usrd to calculate A p r and Ap8.
10
The difference between Ap, and Apz was originally
considered as solids friction, and is given in Tables I11
0
arid IV as apparent Apf.. This value was used in
ISquation 5 t o calculate the apparent values of fa

11

in the tables. C and Re were calculated from


the average values of A ~ J L G , , uB, and (Au) for each
series. For convenience in calculating C and Rr the
expression (0.0O45p,/Dppp) in Equation 16 was grouped
as l / K ? , and D p p p / p o of the Reynolds number function as Kz. Values of K1 and Kz are shown in Table
I1 for each material, based on air at 80" F. and 0.3 inch
.of mercury gage pressure. The carrier air temperature
varied from 74" t o 87" F. and the riser static pressure from 0.1 to 0.8 inch of mercury, but the average values were considered sufficiently accurate for
G and Re. However, ug was based on the measured
temperature and pressure. Experimental values of C
against Re are shown in Figure 10.
T o determine a n order of magnitude of the acceleration effect on the apparent solids friction, the measurements of Ap in the upper and lower halves of the larger
iiser were used t o calculate the mass-average velocity in
the two halves, as shown in Table V. The measured ratio
of Ap8 in the lower half t o t h a t in the upper half was converted t o Ap,/LG, for the two sections, based on the overall Value shown in Table Iv for the corresponding sand
and air velocity. Equation 16 was used graphically to
,hewn

2
Ap,

In. He0

IO

05
Ap,

F~~~~~ 6.

pressure D~~~
against Solids FloC Rate a t
Constant Carrier Air Rates
for Tube 1
35

15

20

25

In. Hz0

Figure 7. Pressure Drop


against Solids Flow Rate a t
Constant Carrier Air Rates
for T u b e 2

MATERIAL A, TUBE NO I
VOL* I 32x1OiFTJ

30 ----

-1

25

7
"

3 20

" 5

!?

15

--

T
IO

2
I

0
A,,

In. Hz0

F,gure
Measurements of
Density of Dispersed Solids in
Tube 1

.{-IO

15

20

25

Ap, In. Ha0

Figure 9. Measurements
of Density of Dispersed
Solids i n Tube 2

1156

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 41, No. 6

relatively simple mathematical treatmcnt. For


these reasons its uie is recommended.
LEGEND
Because of the approximations necessary in
POINT MATERIAL
the analysis of t!ie results, measurements of the
accelerating effect were limited in the present
work to the runs shown in Table Ti. I n a continuation of this problem, one of the phases that
I
\vi11 be studied is the pressure drop in vertical glass
and steel pipes in a section where solids acccleration is negligiblc. With more accurate determinations of the friction loss, it,s mechanism and the
variables which affect it will be better undcrstood.
The experimental values of the drag coefficient, C, plotted against the Reynolds number
based on the slip velocity show remarkably close
agreement with the line for free-falling spheres in
IO'
IO0
IO'
IOP
I oa
IO'
still fluids, Figure 10. The average deviation
of the data for the four sands froin the line for
Figure 10
free-falling spheres is represented by the dashed
line, with the average C I CO(where COis the drag
the loivest air velocity are in fair agreement with the calculated
coefficient for free-falling spheres and C the obscrvrd value, both
free-falling velocities of spheres of the same diameter, shown in
a t the same Reynolds number) equal to 1.47. This deviation
Table 11. I t should be remembered that a nonspherical particle
factor is believed to be due to a combination of the follo-ing:
will have a lover free-falling velocity than a sphere of the same
shape of the particles; surface roughness of the particles; the
volume.
arbitrary expression of the mean particle diameter as the arithThe solids velocities reported in Tables I11 and I\- represent
nietic mean of the smallest and largest scieen openings used:
integrated averages during a n accelerating period; these were
the velocity gradient and turbulence of the carrier gas;
calculated by dividing the mass velocit,y by the average disa hindered settling behavior due to the effect of the particles on
persed solids density.
the flow pattern of the fluid; and the use of an average solids
The data for cracking catalysts arc of liniited value, as is
velocity for cases in w!iic!i particle accelrration was actually
evident in the tables. The low fluidized density of these mateorcurring. Of thcsc, an estimated correction for the departure
rials, compared to t,hat for the sands (30 pounds per cubic
of the particles from a spherical shape brings the C/COratio to a
foot, as compared with 80 to 90 pounds per cubic foot) limited
more uniform value and nearer unity as shown in Table VI.
the available slide valve pressure drop and therefore the maximum solids flow rate a t maxiinuin valve opening. Solids vclocities were high, causing l o r values of p d s . Since the latter value
Table V I . Values of C/COand Shape Factor tor Four Sands
appeared most subject t o experimental error, values of u s calculated froin p d s are not very accurate. The estimated value of
ug is also questionable, since consistent measuremcnts of the
1.22
1 OJ
1.16
h
amount of gas leaking t,hrough the slide valve could not be made
1.20
1.65
1.40
B
1.60
1.25
C
1.28
when the solids in the standpipe w ~ r cin a fluidized condition.
1.46
1.25
1.17
D
However, the data do indicate t,hat with these low-density fine
particles, the relative velocity of the carrier gas and solids is small.
I n view of the data in Table IT,
shon-ing the large effect of
Thc shapc factors were assigned by Leva (9) on the basis CJf
solids acceleration on the apparent solids friction drop, t,he small
the photoinicrographs shoxn in Figure 11. The only evidence
variation of f 8 calculated by assuming no accelerattion, as shown
regarding the velocity gradient and turbulence of the carrier gas
in Tables I11 and IV, a t first appears unexpected. HoiTever, it
is as follows: in two series of runs (materials C and D in the smaller
can be shown that a const,ant fractional increase in 2 ( s per unit
tube) the carrier gas flow, in the absence of solids, was mcll within
of length will result in a constant additive increase in .fa for all
the viscous region. I\-evertheless, the relation between C and Re
values of ut. The second and third serics of Table V, using the
smallest sand, do show t,he same fractional increase (about 15bz,) did not differ from that in cases n-here the gas flow was definitely
turbulent,. This raises a question as to the legitimacy of assumin u s a t velocities of about 10 and 20 feet per second. The first
ing that the pressure drop of the carrier gas, in the presence of
series, with the largest sand, shows only a slightly greater fracsolids, is the same as if the solids w x e absent. The only answer
tional increase (about 2lyG).There is some indication, thereis that deducting this pressure drop from the tota,l pressure
fore, that the fractional increase of us in all runs does not vary
drop, as is done in these calculations, results in values of the drag
greatly. Due to the acceleration effect, the present data offer
coefficient, C, that are consistent and also quit,e near those for
no rigorous justification for the use of the Fanning equation for
free-falling
spheres. The present' lyorlr does not provide incorrelating the actual solids friction drop. However, after
formation regarding any of the other four items.
allowing for the acceleration effect in the three series of runs in
The carbon dioxide run checks alniost exactly the data for air
Table V, the actual frict,ion factor appears t o be practically conin Figure 10, showing that the difference in viscosity and density
stant a t 0.001 within the accuracy of the experimental data.
is accounted for in the correlation. The points for the cracking
The Fanning equation, when used with the solids velocity and discatalysts E and F are only approximations due to the uncerpersed density, is simply a mathematical expression for the statetainty of the calculated slip velocity. Since the slip velocity of a
ment that the particles lose a constant fraction of their kinetic
material of this size and density is so low, an error in calculating
energy,
foot pounds per pound, due to friction b e b e e n the
it has a relatively minor effect on the particle velocity if the carrier
2g
velocity is in the usual operating range of 20 to 30 feet per
particles and the wall of the riser, and between the particles
second, typical in t,he riser of a catalytic cracking unit. Recausc
themselves. Finally, its form in Equation 10 lends itself to

June 1949

INDUSTRIAL AND ENGINEERING CHEMISTRY

A = 25- t o 28-mesh O t t a w a sand

D = 60- t o 80-mesh sea sand

B = 35- t o 48-mesh O t t a w a sand

C = 48- t o 60-mesh sea sand

F=

100- t o 200-mesh rnicrospheroidal catalyst

Figure 11.

1157

100- t o 100-mesh ground catalyst

M a t e r i a l Used

(I scale division = 100 microns)

of the agreement found between thc free-fall line of Figure 10


a n d the data for sand, where the slip velocities were measurable,
it is reasonable t o assume t h a t the free-fall line can safely be
used for the 100 to 200 mesh particles.

'

E F F E C T O F A RANGE O F P A R T I C L E S I Z E S

When the material being transported covers a wide range of


particle sizes, as is desirable in fluidized systems, some average
diameter must be used in calculating the mass-average slip velocity. If it is assumed t h a t interferencc between particles is so
great t h a t all of them are moving a t the same velocity, the diameter of a particle having a n average cross-sectional area would
be the logical choice, since the force on each particle, by the drag
correlation, is proportional t o the product of. ( A u ) ~and A p .
However, i t is more likely t h a t a velocity gradient exists, whereby
the largest particles are moving at the lowest velocity (with the
highest slip). I n this case, the average diameter to use would be
somewhere between the cross-sectional average diameter and the
diameter of the largest particle. By way of example, the effect
of diameter on the free-falling velocity of cracking catalyst in
air at 900" F. and 10 pounds per square inch gage is as follows:
40 mesh (420 microns), 4.8 feet per second; 200 mesh (74
microns), 0.28 feet per second; 5 microns, 0.001 feet per second.
For large risers, where the wall friction is small, the slip velocity
can be assumed t o equal the free-falling velocity. Some sort of
average free-falling velocity presumably would apply for the
usual case of a mixture of sizes, b u t the present work with materials of narrow size range provides no information on this point.
S A M P L E C A L C U L A T l ONS

The method of calculation of the values listed in Tables I V


and V will be illustrated for material A in the 0.532-inch riser

with a carrier air velocity of 29.5 feet per second. Equations


22 through 25 will be applied also t o calculate the change of
solids velocity and pressure drop with vertical distance during
acceleration of the solids from zero velocity to the equilibrium
value.
A. Table IV:
At 1.10 inch H20 total Ap,

w s = 31.6 pounds/hour from Figure 7

.If8= 1.56 grams from Figure 9

Cross sectional area of riser = 0.00155 square foot

G, =

31'6
= 5.66 pounds/(square foot)(second)
0.00155 X 3600
\701ume of riser between by-pass valves = 6.86 X 10-3 cubic foot

'Oo0 = 0.501 pound/cubic foot


=
454 X 6.86
Bverage us = G8/pda = 5.66/0.501 = 11.3 feet/second
Distance between pressure taps = 40.1 inches
Solids static head = A p , = 40.1 X pds/62.3 = 40.1 X 0.501,'
62.3 = 0.32 inch of H20
Air friction ressure drop a t 29.4 feet/second = A p f , = 0.56
inch of Hs8 from Figure 7
Solids pressure drop
A p s = A p - A p f s = 1.10 - 0.56 = 0.54
inch of HzO
Apparent solids friction drop = A p B - A p z = 0.54 - 0.32 =
0.22 inch of H20
Pds

By Equation 7,

APlQD? =
0.22 X 32.2 X 0.0444
= o.oo38
0.385LG,u8
0.385 X 3.34 X 5.66 X 11.3

slip velocity

Au = ug - us = 29.5

- 11.3 =

By Equation 16,

= APs

DPPP

Ua

LG, 0 . 0 0 4 5 ~( ~A u ) ~

18.2 feet/second

INDUSTRIAL A N D E N G I N E E R I N G CHEMISTRY

1158
D B __
pp
=

0.0045~~

Vol. 41, No. 6

808 for material A in air ( K , in Table 11)

0 5 4 X 808 X 11.3
3.34 X 5.66 X (18.2)2

( J - L- _ _ _ ~

Re

079

n__
D (Au),,
MU

= 10.2 for material A in air ( K 2 in Tablc 11)


Po

Re = 10.2 X 18.2

186

These values of C arid RP fall evartlp on the frw-fall iiI1cl of Figure


10.
R. Acceleration measureinellti, Tahlr V :
I n measuring the effect of solids acceleration 011 the pressure
drop across the two halves of the riser, Ape (Iowcr)//Ip, (upper)
was found to be constant at 1.47 over a range of solids flow rates
for material -4at an average ug of 29.4 (Table V). This ratio
and the over-all value of Ap,/LG, of 0.0286 from Table I T T were
used t o calculate apg/LG, for the lower and upper sections a t
0.0232 and 0.0340, respectively. By Equation 16, Aps/LG8 was
plotted against us,using the free-fall line of Figure 10 to relaw C
t o ( A u ) since the average C and Re fell on this line. This plot
showed us t o be 10.3 at Ap,/LG, = 0.0340 and 12.5 at 0.0232.
It was assumed t h a t these velocities occurred at the mid-points
of the two halves.
The acceleration component of Equation 25

US,

Ft./Sec.

Plot of Acceleration

Figure 12.

Equations
(Sample calculation)

dashed straight line in Figure 12, r h i c h permits a mathc.niat,ical


integrat,ion. The straight line must coincide as nearly as possi1,lc
with the curve a t low values of

(1,

u. is a largc:

where

t1u11d)c~l~.

At high values of a, tho diffcrencc bct,wcen the two lincs


niinor effect on theintegrat,ion brcauao here u / n is small.
straight line,

Solving for the contribution of acceleration to ,fa,

a;, -

(1

(I

liah

(tit!

(2(ij

)i

11,

wlierc U A is the: j/-int,c,


pl, = N!: ur = cquilibriuni u s = 14.2s.
Substituting t,liis expi io11 for (I in l?qiiat:ion 24 and i t i t v g i ~ a t i i ~ g
ht,tn-cen
= 0 nnct w1 = i i r rewilts i l l
1,

I&

lie

ii,

1i:rainples of
I&

= 0

I, = 0

For the same conditions as above (material A , 0.532-i tirh risw,


average ? i o = 29.4 feet /second)

-AB= 0.206 recipi,oca~tcct


4P,D,

aPSUlxlej*

0.001

D,.
1s

29.4
23.4

L cdculatcttl
2

I/,$

( N i l )

>

i t r 1 his mtiiiii(~rtire tis follons:


10

0.03

1 .29

7 . 5 2 feet

This luiiction is shown in Liiguro 12 as thc curved lint: I,! rll(~Jretically, the eyuilibrium velocity is never reached. H ~ ~ Y C Y W ,
SSTl of the final velocity is attained in this casc a t 7 . 5 f e ~ abovct
t
the t,hcoretical aero I c ~ ~ e l I. n the prcsent, casc, in which t11(%
vertical section is preceded by n bend, the zero levcl c:annoi. I)c
ascribed any physical significance until more is learned n t w u t llrc
niechanism of flow of the solids in curvc~Isections.

0.0454 ,eciproc:al feet

u !viis calculated over a range of


from tcs = 0
ing t o a = 0. Examples arc given below:

0
0
14

111
lJg

(1

to u s corlespi)iiti-

0.65
116
0
84
0.72
81 ..?
1.6
1 7 , .?I
167
0,87
1 2 ,a
8.9
1.1
I l e = Kz(Au,) = 10.2 ( A L L ) .
F r o m the free-fall line i n Figiirc 10, sirlcr riir? o h a e r i e d aielaae R e a n d C
300

239

15.4

fit thiq line.

Values of a 5 0 obtained arc plottcd agaiiist n, in Figurca I 2


(the curvcd solid line). To calculate, 1, as R function of u,,the
folloiring integration m-aa perfozined
(24)

This can bc don? gi,aphically by plotting exact values of u,/a


against I / ,
FTowrver, littlc (war is ~ntroduced by using the

L = Distance, Ft.

Figure 13. Pressure Drop and


Solids Velocity against Distance
during Acceleration
(Sample calculation)

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

June 1949

1Squation 25 was used to calculate the differential pressure


drop as a function of us.

-1constant value of 0.001 was again assumed for f 8 .

x, (2)

0.1925 0.100

+ 0.0014 X

At us = 10,

1. What is the catalyst flow rate if the effective average


particle diameter is (a) 50 microns; ( b ) 100 microns? Assume
th&tthe solids friction drop is negligible.
2. Show t h a t the solids friction drop is a negligible fraction
of the total pressure drop. The value for f s for cracking catalyst
in a brick-lined pipe is not known, but use values of 0.001 and
0.004 t o cover the possible range.

10 4SOLUTION

a&)

= 0.0361

1.

Keglecting solids friction, Equation 10 becomes

us = 2
- - ( d p / d L ) = 0.312

G,

12

14.1

0.124

0.02A2 0.0181

14

0.0178

This relation is plottrd also in Figure 12 as the curve marked


dl).

The values o

as(dp/clL) and

itgairiat L as abscissa.

us are rcplotted in Figure 13

I n this form, the area under the d p

c'urvc' (line A) represents total

between
G,

US

0.1925 or G

Apa

mi = -

Similarly, other values are as follows:

1159

APsua
0.1925 L

It has been shown t h a t if the solids friction is negligible, the slip


velocity, Au, will be equal t o the free falling velocity of the particles.

(a)D,

= 50 microns = 1.64 X

feet.

For particles of this diameter, free-fall is in the viscous region


where C = 24/Re (Figure 10). With no solids friction, the forcr
on one particle will be equal to its weight or (?/6)D;(pp - p q ) .
Substituting these values of C and F in Equation 11 results 111
Stokes' law

any two values of

A considerable distance must be traveled before the accelerating ctffect becomes insignificant.
This theoretical approach was substantiated by the measurements of Ap in the two halves of the riser (Table V). The distancc between the outer pressure taps, length ab, was positioned by
/..

so t h a t the integrated average -"


'- that is,
LG,
the area under the curve between the abscissa values of a and
t i ial-and-error

b -corresponded t o the measured over-all value of 0.0286.


The abscissa at d' then corresponds t o the center pressure tap.
The ordinates of cd and ef are the measured values of Ap,/LG,
in the lower and upper halves, respectively. The agreement
between these measured values and the integrated average values
for the two halves (obtained by measuring areas) is excellent.
L\lso, the integrated average solids velocity from curve B in
Figure 13 over the length ab agrees with the measured average
us of 11.3. These two agreements confirm the above method of
iriterpretation of the observed data.
Curve A , Figure 13, also indicates that the distance corresponding t o a hypothetical zero velocity (u, = 0, L = 0) was
immediately above the lower by-pass valve, which was 9 inches
below the bottom pressure tap. This probably means t h a t constdwable interference t o solids flow occurred due to the combined
;tction of the bend and the by-pass valve.
The pressure drop required to accelerate the solids cannot be
calculated by multiplying the increase in kinetic energy of the
particles (foot pounds per pound) by the solid-to-gas flow ratio
(pounds solids per pound of gas) t o yield foot pounds per pound
of gas, because such a calculation assumes 100% energy transfer
troni the gas t o the solids. Actually, the efficiency of energy
tiansfer is at all times equal t o us/uB. For any slip velocity,
there exists a definite force F , exerted by the gas on an individual
particle. I n one second the mechanical energy lost by the gas
= F p X u,, while the mechanical work done on the particle is only
F,, x 2L".
APPLICATION

111 a fluid catalytic cracking unit spent catalyst riser, 6 feet in


iriside diameter, air at 900' F., a n average pressure of 10 pounds
J ) C ~square inch gage, and a linear velocity of 25 feet per second is
used to transport catalyst from the reactor slide valve t o the
rrgenerator. A differential pressure recorder connected to taps
across the upper half of the 100-foot vertical duct (where solids
ncceleration is negligible) shows a reading of 0.25 inch of water
per foot. With the same air flow and no catalyst circulation, the
meter indicates a zero pressure drop.

With p p equal to 61 pounds per cubic foot (Table 11) and pu


0.032 X 0.000672 pound per (foot)(second)

Au = 0.14 foot per second


Ua =

ug -

471 =

25.0 - 0.1 = 24.9

From Rquation 26

0,=

0'2:,:2y'9

= 32.4 pounds per (square foot)(secontl)

The catalyst flow in tons/hour

32.4 X 3600 X 0.785 X 362000

1650.
( b ) D, = 100 microns = 3.28 X

feet

I n air a t these conditions, a particle this size still conforms to


Stokes' law during frec fall, so
AIL = (100/50)2 X 0.14 = 0.56 foot per second

(This corresponds t o Re = 0.42, which is well in the viscous


range, Figure 10, showing that Stokes' law is applicable.)
U.

= 25.0

- 0.6

= 24.4

0.25 X 24.4 = 31.7


G, = ,
0.1925
I n tous per hour, catalyst flow = 31.7 X 1.8 X 28.3 = 1615.
Thus, the value used for the particle diameter is of minor importance in this calculation when the air velocity is high compared to the slip velocity.
2. Effect of solids friction:

LG,=

0.192511/u8

+ (2f,lgD,)u,l

(10)

If the solids friction is appreciable, the slip velocity will be higher


than t h a t calculated for free-fall. Arbitrarily using a value of 24
for ul, the relative values of static and friction pressure drops can
be calculated over'a range offs
f s = 0.001
0.004
1
Relative static A p = U,

Relative friction Ap

Sf&,
= -gDr

Total

Friction A p a s a p e r centof thetotal

0.0416
0.0003

0.0416
0.0011

0.0419

0.0427

0.7%

2.6%

This shows t h a t for all practical purposes, the solids friction


drop can be neglected in this articular problem and Equation 26
can be used directly with a s i p velocity equal t o the free-falling
velocity.

INDUSTRIAL AND ENGINEERING CHEMISTRY

1160

This problem illustrates how, without knowing exactly what


effective particle diameter to use for a mixture of sizes and the
value of the solids friction factor, f8, a relation between solids
flow rate and pressure drop can be estimated. This value of
pressure drop holds only in t h a t part of the vertical riser where
particle acceleration is negligible. It should be emphasized t h a t
in a complete system the pressure drops due t o initial particle
acceleration, bends, and acceleration after bends can be a major
portion of the total, indicating the need for considerable experimental work on these phases of the problem.
CONCLUSION

It has been shoivn t h a t in a vertical riser through which solid


particles are transported by a stream of gas, the total pressure
drop can be considered the sum of that due to the gas flow alone,
as though no solids n-ere present, and a solids pressure drop,
The latter consists of a solids static head, a solids friction loss
due to contact b e h e e n the particles and the pipe wall, and, for a
considerable distance above the point where the solids start t o
move in a vertical direction, an acceleration pressure drop.
The solids static head can bc expressed as the vertical distance
times the dispersed solids density; the latter is related to the
solids mass and linear velocities. For the friction loss component, the Fanning equation, used with the solids linear velocity
and dispersed solids density, is recommended. Due to a relatively large amount of acceleration pressure drop in the present
experiments, the solids friction loss could not be measured independently. However, after allowing for the acceleration effect
the order of magnitude of the Fanning friction factor for sands in
glass tubing is about 0.001.
The pressure drop t h a t is required to produce acceleration
of the particles in the lower section of a vertical riser was shown by
measurements and by a mathematical analysis to be a significant
portion of the total. An extremely high pressure drop is found
to be possible at the point of introduction of the solids into the
gas stream. Equations were developed for calculating the
variation of pressure drop and solids velocity with distance in the
accelerating section.
At any time, the solids velocity corresponds to a state of dynamic equilibrium, in which the rctarding forces of gravity, friction, and inertia must be exactly equaled by the force exerted by
the gas OII the particles. The latter force was shown to be related
to the slip velocity by thc drag correlation for free-falling particles in still fluids. This is a completely unexpected situation,
since the turbulent flow conditions existing in the transport of
solid particles by a gas bear no resemblance to those for the free
fall of individual particles in a still fluid. Deviation from the
relation for free-falling spheres is believed mostly due to the irregular shape of the particles; this can be corrected for by a n
appropriate shape factor.
To extend the theory of particle transport, further experimental
M-ork is needed. The following are some of the many interesting
problems that have come to mind during the course of the present
work:

D,
D,
F

VoL 41, No. 6

= diameter of a part cle, feet

diameter of the riser, feet


force, pounds
F,
force on one particle, pounds
fs
Fanning friction factor for the solid particles, dimensionless
G, = mass velocity of solids, pounds per second per square foot
g
= acceleration of gravity, 32.2 feet per second per second
K1 = grouping of the constant and solid and air properties in t h e
drag equation (Equation 16), equal t o D,p,/0.0O45pv,
inches of water times square seconds per pound
K Z = grouping of the solids and air properties in Reynolds number, = Dppy/,uo,seconds per foot
L = vertical distance in riser, feet
Ma = observed weight of solids between the riser by-,pass valves,
grams
A = number of solid particles in L feet of riser
Re = Reynolds number based on the particle diameter and slip
velocity = Dp(AILj p o / p g , dimensionless
u0 = gas vclocit,y, feet per second
tia
= mass-average velocity of solids particles, feet per second
ut
= terminal free-falling velocity of a part,icle, feet, per second
w s = flow rate of solids, pounds per hour
A p = total observed pressure drop across L feet, inches of water
A p f , = pressure drop in L feet due to gas friction, inches of water
Ap/,= pressure drop in L feet due t o solids friction, inches of
wat,er
A p , = total solids pressure drop in L feet = ApJa
Ap,,
inches of water
ApZ = solids &tic pressure drop in L feet, inches of water
Au = slip velocity = tio - 2 1 , feet per second
X = particle area-volume shape factor, dimensionless
po = gas viscosity, pounds per foot second
pda
= dispersed solids densit,?, pounds of dispersed solids per
cubic foot
po
= gas density, pounds per cuhic foot,
pp
= particle density, poufids per cubic foot,
=
=
=
=

LITERATURE CITED

(1) Chatiey, H., Engineering, 149, 230 (1940).


Chemistru &: Industru, 44, 207 (1925).
(2) Cramp, W,,

(3) Dalla Valle, J. M., Micromeritics, 1st ed., New York, Pitman

Publishing Corp., 1940.


(4) Daniels, L. S., Petroleum Re.fi?zer, 25, 435 (1946).
(5) Hudson, W.G., Chem. & X e t . Eng., 51, 147 (1944).
(6) Hudson, W. G., Conveyors and Related Equipment, X e F
York, John Wiley & Sons, 1944.
(7) Jennings, M.,
Engineering, 150, 361 (1940).
(8) Lapple, C. E., and Shepherd, C. B., IBD.E N GCHEM.,
.
32, 606
(1940).
(9) Leva, bl.,personal communication.
(10) Perry, J. H., Chemical Engineeis Handbook, 2 n d ed., p. 807,
New York, McGraw-Hill Book Co.. 1941.
(11) I b i d . , p . 1852.
(12) Vogt, E. G., and White, R. R., IND.
ENG.CHEV.,40, 1731 (19481.
(13) Webb, G . M., Petroleum Processing, 2, 397 (1947),
RECEIVED
February 2 5 , 1919.

Effect of a mixture of particle sizes


Horizontal and inclined pipes
Friction losses in commercial piping
Losses in bends
Effect of a bend on the acceleration drop immediately following
it
Losses a t the point of entrance of the solids into the gas stream
NOMENCLATURE

A,
A,
C
CO

= acceleration of particles, feet per second per second


= projected area of a particle, square feet
= cross-sectional area of the riser, square feet
= drag coefficient = 2Fpg/p,( A U ) ~ A dimensionless
~,
= drag coefficient for free-falling spheres from Figure

dimensionless

COURTESY

10,

STANDARD O l L COWPANY

(NEW J E R S E Y )

Fluid Catalyst being Drained f r o m Petrol e u m Cracker during Plant Shutdown

Você também pode gostar