ExxonMobil Proprietary

FLUID SOLIDS Section Page

VI-E 1 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

Changes shown by ➧
CONTENTS

Section Page

SCOPE ............................................................................................................................................................3

REFERENCES ................................................................................................................................................3

ADDITIONAL LITERATURE ...........................................................................................................................3

FLUIDIZATION ................................................................................................................................................3
DESCRIPTION OF FLUIDIZATION ........................................................................................................3
SUPERFICIAL GAS VELOCITY..............................................................................................................4
Minimum Fluidization Velocity ..............................................................................................................4
Particle Terminal Velocity .....................................................................................................................4
PROPERTIES OF BUBBLES..................................................................................................................5
Bubble Velocities ..................................................................................................................................5
Bubble Size ..........................................................................................................................................6
ENTRAINMENT ......................................................................................................................................7
Transport Disengaging Height (TDH) ...................................................................................................7
Entrainment Correlation........................................................................................................................8

FLUID BED PROPERTIES, PERFORMANCE................................................................................................9

FLUID BED DENSITY .............................................................................................................................9
BED VOIDAGE........................................................................................................................................9
Dilute Phase Density ............................................................................................................................9
Dilute Phase Holdup...........................................................................................................................10
GAS / SOLID CONTACTING ................................................................................................................10
Bubble Phase .....................................................................................................................................10
Emulsion Phase .................................................................................................................................11
Interchange Between Bubble And Emulsion Gas...............................................................................11
Gas Dispersion Coefficients ...............................................................................................................11
SOLIDS MIXING AND SEGREGATION................................................................................................12
Solids Mixing ......................................................................................................................................12
Gulf Streaming ...................................................................................................................................12
Segregation ........................................................................................................................................12
CHARACTERIZATION OF GAS, SOLIDS BACKMIXING .....................................................................13
GAS, SOLIDS FLUID DYNAMIC CALCULATIONS ..............................................................................13

SOLID PROPERTIES....................................................................................................................................14
PARTICLE SIZE....................................................................................................................................14
Size Distribution .................................................................................................................................14
Measurement Techniques ..................................................................................................................15
SOLID DENSITY ...................................................................................................................................18
Skeletal Density..................................................................................................................................18
Particle Density ..................................................................................................................................18
Bulk Density .......................................................................................................................................19
Pore Volume.......................................................................................................................................19

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 2 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

CONTENTS (Cont)

Section Page
SOLIDS FLOW PROPERTIES ..............................................................................................................19
Angle of Repose .................................................................................................................................19
Angle of Internal Friction.....................................................................................................................20
Angle of Slide .....................................................................................................................................20
ATTRITABILITY.....................................................................................................................................20
EROSION ..............................................................................................................................................20
Types of Erosion.................................................................................................................................21
Factors Affecting Erosion....................................................................................................................21
Current State of Technology, Recommendations ...............................................................................21

SAMPLE CALCULATIONS ...........................................................................................................................22

MINIMUM FLUIDIZATION VELOCITY, VOLUME, SURFACE MEAN PARTICLE DIAMETER .............22
TERMINAL VELOCITY..........................................................................................................................23
BUBBLE DIAMETER, RISE VELOCITY................................................................................................24
TRANSPORT DISENGAGING HEIGHT................................................................................................24
DENSITY, VOIDAGE.............................................................................................................................25

NOMENCLATURE.........................................................................................................................................26

TABLES
Table 1 Heywood Shape Factors for Representative Solids.........................................................28
Table 2 Comparison of Various Standard Sieve Series ................................................................29
Table 3 Representative Fluidized Solids Properties FCC Catalyst Size Distribution....................31

FIGURES
Figure 1 Terminal Settling Velocity, Uto or Utcl ...............................................................................34
Figure 2 Effective Cluster Diameter Based on Vertical Conveying Data ........................................35
Figure 3 Series of Tank Models for Fluid Solids Process...............................................................36
Figure 4 Log-Normal Probability Particle Size Distribution.............................................................37
Figure 5 Effective Range of Particle Size Measurement Techniques.............................................38
Figure 6 Angle of Repose for Granular Solids................................................................................39
Figure 7 Angle of Internal Friction ..................................................................................................39
Figure 8 Ductile and Brittle Erosion................................................................................................40

Revision Memo
12/02 Corrections and minor updates.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 3 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

SCOPE
This section describes the basic properties of fluid beds and particles. It provides the background information needed in the
design of fluid solids systems. The information is of a general nature and is applicable to all fluid solids processes used within
Exxon. However, when dealing with a known process, data and practices based on commercial plant experience should be used
when available.

REFERENCES
1. Wen, C. Y. and Yu, Y. H., AlChE Journal 12, 610 (1966).
2. Rossetti, S. J., Design and Troubleshooting of Pneumatic and Other Gas Conveying Systems, EE.3E.76, February 1976.
3. Coulaloglou, C. A., An Improved Correlation for Bubble Sizes in Fluidized Beds, EE.48E.77, March 1977.
4. Werther, J., Bubble Growth in Large Diameter Fluidized Beds, Fluidization Technology Vol. 1, Hemisphere Publishing Corp.,
Washington, D.C. (1976).
5. Shabaker, R. H., An Improved Correlation for Entrainment from Fluidized Beds, EE.123E.77, December, 1977.
6. Halow, J. S., Dilute Phase Density Above Fluidized Beds, EE.34LDL.70, July 1970.
7. Kunii, D., Levenspiel O., Fluidization Engineering, Robert E. Krieger Publishing Co., Huntington, New York (1977).
8. Levenspiel, O., Chemical Reaction Engineering, John Wiley & Sons, Inc., New York (1972).
9. Manual on Test Sieving Methods, ASTM Special Technical Publication 44.
10. Byers, R. L., Stack Sampling Guide, EE.119.76, December 20, 1976.
11. Halow, J. S., Erosivity of Cracking Catalyst, EE.43E.77, March 1977.
12. Woodward, J. L., Erosivity of Cracking Catalyst to Commercial Refractory Lining Material, EE.38E.86, March 1986.
➧ 13. Skouby, D. C., Sharma, A. K., Improved Fluidized Bed Entrainment Correlation, EE.701E.2002, December 2002.

ADDITIONAL LITERATURE
1. Zenz, F. A. and D. F. Othmer, Fluidization and Fluid-Particle Systems, Rheinhold Chemical Engineering Series, New York
(1960).
2. Matsen, J. M., Rossetti, S. J., and Halow, J. S., Fluidized Beds and Gas-Particle Systems, Encyclopedia of Chemical
Processing and Design, McKetta, J. J., Editor. 23, 80 - 196 Marcel Dekker, New York (1985).
3. Geldart, D., Ed., Gas Fluidization Technology, John Wiley & Sons, New York (1986).
4. Davidson, J. F., Clift, R., and Harrison, D., Fluidization, Academic Press, London (1985).

FLUIDIZATION

DESCRIPTION OF FLUIDIZATION
When a fluid flows upward through a bed of solid particles, the pressure difference across the bed increases approximately
linearly with velocity. At a sufficiently high velocity (minimum fluidization velocity, Umf) the drag force on the particles becomes
equal to their weight and the particles become neutrally buoyant. The particles are said to be fluidized and the particle bed takes
on characteristics of a liquid.
As velocity is increased beyond the minimum fluidization velocity, the bed expands, and under ideal conditions, the pressure
difference across the bed remains nearly constant and equal to the weight of the bed per unit cross-sectional area. Above the
minimum fluidization velocity, the bed can expand in two ways. In particulate fluidization, beds expand uniformly, with the
particles maintaining a somewhat regular spacing as they move apart. Fixed bed laws for pressure drop (e.g., the Carman-
Kozeny or the Ergun equation) are applicable, with the provision that bed voidage, ε increases with fluid velocity in order to
maintain constant pressure drop, rather than being independent of velocity as for a fixed bed.
In aggregative or bubbling bed fluidization, the bed is non-uniform. Most of the excess gas passes through the bed in fast-
moving bubbles (bubble phase) which are nearly empty of particles, while the rest of the bed, called the emulsion phase, remains
close to minimum fluidization conditions. Aggregative fluidization may become quite violent at high fluid velocities, and the bed
structure and fluid flow patterns are much more complex than in particulate fluidization. Liquid fluidized systems generally display
particulate behavior while gas fluidized systems are aggregative. Only gas-fluidized systems have achieved commercial
significance, and particulate systems are of mainly academic interest.
The major feature of an aggregatively fluidized bed is the existence of much of the fluidizing gas in the form of large bubbles.
Properties of these bubbles help define different types of gas-solid fluidization.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 4 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

SUPERFICIAL GAS VELOCITY

The performance of a fluidized bed is determined to a large degree by the velocity of the upward moving gas phase. The
measure of the gas velocity normally used is the superficial gas velocity. This is the velocity the gas would have if it were rising
up the fluidized bed vessel, empty of solids, but containing internals such as diplegs, overflow wells, etc., at operating
temperature and pressure. Corrections for internals may not have been included in previous definitions of superficial gas
velocity.

Minimum Fluidization Velocity

Minimum fluidization velocity, Umf, is the superficial velocity at which particles first become suspended by the flowing fluid. Umf
can be calculated theoretically by selecting one of the several available expressions for pressure drop through a packed bed and
determining the fluid velocity at which pressure drop balances the weight of the particles. Empirical expressions for Umf are more
➧ reliable and are more generally used. Minimum fluidization velocity is calculated by PEGASYS cyclone module (see Section VI-
C, Appendix A) or by the following equation which was used to correlate 284 data points over a particle Reynolds Number, Rep,
range of 0.001 to 4000 with a standard deviation of ± 34% (Reference 1).
1/ 2
dp Umf ρ g é 0.0408 dp3 ρ g (ρp − ρ g ) g ù
= ê(33.7)2 + ú − 33.7 Eq. (1)
µ ê µ 2 ú
ë û

dp Umf ρ g
Rep =
µ
where: dp = The volume surface mean particle diameter (ft), (m) [see Eq. (36)]
Umf = The superficial minimum fluidization gas velocity (ft/s), (m/s)
ρg = The gas density (lbs/ft3), (kg/m3)
ρp = The solid particle density (lbs/ft3), (kg/m3) (see SOLID PROPERTIES)
µ = The gas viscosity (lbs (m)/ft/s), (kg/m/s)
➧ g = The gravitational acceleration (32.2 ft/s2), (9.81 m/s2)
For smaller particles, the equation reduces to:
dp2 (ρp − ρ g ) g
Umf = for Rep < 20 ( ft / s), (m / s) Eq. (2)
1650 µ

and for large particles:

1/ 2
é dp2 (ρp − ρ g ) g ù
Umf = ê ú for Rep > 1000 ( ft / s), (m / s) Eq. (3)
ê 24.5 ρ g ú
ë û

A bed with mixed particle size cannot be operated stably for long at minimum fluidization velocity, because the largest particles
may settle out to form a packed, unfluidized zone at the bottom of the bed. If such settling is unacceptable, then the bed must be
operated at the velocity above the minimum fluidization velocity for the largest particles.

Particle Terminal Velocity

At relatively low gas velocities, some solids are ejected from the surface of a bubbling bed, but readily fall back into the bed. At
higher velocities, some of the solids are entrained and transported out of the fluid bed vessel with the gas. Under these
conditions, two phases are present: the dense bubbling bed and the dilute entrained flow phase above it. As the gas velocity is
increased, the character of the bubbling bed changes, solids motion becomes more agitated and the interface between the dense
and dilute phases becomes less distinct. At some higher velocity, no dense phase is discernible and solids are transported
rapidly through the vessel in a dilute phase (vertical pneumatic transport). A steady state dense phase fluidized bed can be
operated at a gas velocity in excess of the terminal or free fall velocity of bed particles only if the entrained particles are captured
and returned to the bed.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 5 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

The single spherical particle terminal (settling) velocity, Uto, can be determined from Stoke's law:
dp2 (ρp − ρ g ) g
Uto = ( ft / s), (m / s) Eq. (4)
18µ

dp Uto ρg
at values of the particle Reynolds number, , less than 1.0. The particle density, ρp is defined under SOLID
µ
PROPERTIES. The particle diameter dp, is the weight median size (see SOLID PROPERTIES).
At higher Reynolds numbers, Uto can be obtained by using Figure 1. For numerical calculations, Figure 1 is represented as:
1 24 0.69
= + Eq. (5)
Y 2 X
X

➧ where: X and Y refer to the axes in Figure 1.

Fine particles at appreciable concentration have settling velocities somewhat higher than the settling velocities for single
particles. This is due to forces which cause the particles to behave as clusters or aggregates. The effective diameter of these
clusters, expressed as multiples of the single particle diameter, is related to the single particle terminal velocity in Figure 2. The
settling velocity of a cluster is calculated as the terminal velocity of a particle with a diameter equal to the effective cluster
diameter. For large particles, this cluster settling velocity approximates that for a single particle. The cluster terminal velocity is
obtained from Figure 1 using values of n, the cluster index, obtained from Figure 2. Data given in Figure 2 were obtained under
vertical pneumatic conveying conditions in a 3 in. (75 mm) pipe (Reference 2). The effect of vessel diameter on settling velocity
appears to be small and is not included in these calculations.
If the terminal velocity, Ut, is exceeded, particles will be carried out of the bed, and the bed will gradually disappear if the
entrained particles are not returned to be bed or new particles added. If particles are returned or added, however, the bed can
retain its identity far above Ut. Thus, although Ut may place a practical limit on simple batch laboratory units, many commercial
units operate at much higher velocities. For units which are fitted with two stages of cyclones housed within the top of the reactor
vessel, maximum design superficial bed velocity is limited by the cyclone capacity which can be installed in the vessels and not
by entrainment rate.
If superficial velocities are increased further, the fluid bed will pass through a transition zone between the dense phase and dilute
phase regimes. The transition zone is characterized by pressure surges and unsteady solids flow. Operation in the transition
zone should be avoided or minimized to the extent possible. The velocity at the transition point is called the choking velocity. At
gas velocities above the choking velocity, a dense bed will not be present and operation will be entirely in the dilute phase flow
regime. The choking velocity can be estimated by first calculating the entrainment for a series of gas superficial velocities and for
the height of the vessel or riser in question [Eq. (15)]. The gas velocity which yields an entrainment rate equal to the solids
circulation rate is the choking velocity.
Commercial beds tend to be designed for gas velocities limited by some upper limit that is characteristic of the particular process
of interest. If internal cyclones are used, the limit is as noted above. If entrained particles are excessively sticky or fouling in
nature, low velocity may be necessary in order to reduce entrainment and particle deposition above the bed. If no cyclones are
to be used, then the velocity and entrainment rate must be selected so as to give tolerable particle losses.

PROPERTIES OF BUBBLES
Fluidized bed properties and performance are, in many cases, related to the size and rise velocity of the bubbles in a bubbling
dense bed. Correlations have been developed relating bubble properties to operating conditions. These correlations are then
used to predict fluidized bed properties and performance.
Gas bubbles in fluid beds have many of the properties of gas bubbles in low viscosity liquids. Fluid bed bubbles have the same
spherical cap shape and flattened or concave bottom as those in liquids.

Bubble Velocities
• Single Bubble Rise Velocity - Rise velocity, UB, of single bubbles in incipiently fluidized beds is given by the following
relationship:
UB = 0.71 gDB ( ft / s), (m / s) Eq. (6)

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 6 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

When the frontal diameter of the bubble DB becomes greater than 0.3 times the bed diameter D, the bubble is called a slug,
and the relationship becomes:
UB = 0.35 gD ( ft / s), (m / s) Eq. (7)

This is also applicable to liquid systems and is theoretically derivable. A slug, instead of being centered (axisymmetric), may
rise along one side of the bed, behaving as if in a bed of twice the actual diameter. This equation then becomes:
UB = 0.49 gD ( ft / s), (m / s) Eq. (8)

• Bubble Velocity During Continuous Fluidization - When a bed is fluidized by a substantial flow of gas, bubbles are
generated continuously, and the bubble rise velocity is greater than when only a single bubble is present. The bubble rise
velocity relative to a stationary observer, UA, is given by:
UA = UB + U − Umf ( ft / s), (m / s) Eq. (9)

where: U = The superficial gas velocity

The increase in bubble rise velocity over the case of a single bubble is U – Umf which is the velocity at which the bed of
fluidized solids itself is being moved upward by the incoming gas.
This is for an idealized case in which the time average gas flow is uniform across the bed cross section, in which there are
no persistent solids circulation patterns, and in which negligible bubble coalescence occurs. In many large fluid beds,
coalescence occurs, and gulf stream flow exists in which gas flow is channeled through the center of the bed and solids
circulate down at the walls and up in the center. In the case of slug flow, however, the expression for UA has been proven to
be generally applicable.

Bubble Size
Bubbles formed at the grid will coalesce and grow as they rise up through the fluidized bed until a stable size is reached. A
stable size is reached when the larger bubbles begin to break into smaller bubbles. One theory of bubble breakage is based on
the tendency of bubbles to entrain solids as the bubbles move up the bed. As bubbles grow, their velocities also increase until
the bubble velocities approach terminal velocity of the particles. At that point, solids will accumulate in the larger bubbles causing
the bubbles to split into smaller bubbles. An equation giving the mean stable bubble size is based on this theory (Reference 3).
➧ The mean stable bubble size is calculated by PEGASYS cyclone module (see Section VI-C, Appendix A) or by the following
equations:
0.666
é (ρp − ρg )2 ù
DB = 0.03 ê ú (dp) 1.46 ( ft ) Eq. (10c)
ê ρg µ ú
ë û

0.666
é (ρp − ρg )2 ù
DB = 0.011 ê ú (dp) 1.46 (m) Eq. (10m)
ê ρg µ ú
ë û

where: DB = The mean stable bubble size (ft), (m)

dp = The volume surface particle diameter [Eq. (36) (ft), (m)]
ρp = The particle density (lbs/ft3), kg/m3)
ρg = The gas density (lbs/ft3), (kg/m3)
µ = The gas viscosity (lbs/ft/s), (kg/m/s)
This correlation is applicable over a cluster Reynolds Number, Recl, between 0.4 and 500, where Recl is defined as:
dcl Utcl ρ g
Re cl = Eq. (11)
µ

where: dcl = The cluster diameter calculated for Eq. (11) only as:
dcl = 0.264 (dp)0.73 (ft)
dcl = 0.192 (dp) 0.73 (m)
Utcl = The cluster terminal velocity from Figure 1 where n for this application is the ratio dc / dp
from Figure 2 (ft/s), (m/s)

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 7 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

Bubbles achieve the mean stable size fairly close to the grid. In those cases where deep beds are used, the bubble growth
region is not important and the mean stable size can be used for fluidization correlations. In shallow bed reactors, or in those
cases where rapid reactions occur in the vicinity of the grid, the rate at which the bubbles grow may be important. The bubble
size can be calculated at various heights above the grid from the empirical Eq. (4).
DB = 2.8 x 10 −2 (1 + 8.29 (U − Umf ))1/ 3 (1 + 2.08 h) 1.21 ( ft ) Eq. (12c)

DB = 8.5 x 10 −3 (1 + 27.2 (U − Umf )) 1 / 3 (1 + 6.82 h) 1.21 (m) Eq. (12m)

where: DB = The bubble size (ft), (m)

U = Superficial gas velocity (ft/s), (m/s)
Umf = Superficial gas velocity at minimum fluidization conditions (ft/s), (m/s)
h = The height above the grid (ft), (m)
The height at which the stable bubble size is reached can be estimated by equating Eqs. (10) and (12) and solving for h.

ENTRAINMENT
Solids entrainment is defined as the weight of solids per volume of gas flowing past a point in the dilute phase region above a
dense fluidized bed. The entrainment is usually calculated at the vessel outlet or at the inlet to internal cyclones. Entrainment
differs from dilute phase density, discussed later, which is the concentration of solids in a given region above the dense bed,
expressed as weight of solids per volume of dilute phase. The entrainment decreases with height above the dense bed, also
referred to as outage, up to a point above which the entrainment becomes constant. The value of the outage at which the
entrainment becomes constant is called the Transport Disengaging Height or TDH. TDH is used in entrainment and dilute phase
holdup calculations.

Transport Disengaging Height (TDH)

A bubble breaking at the surface of a fluidized bed creates a high velocity jet which ejects solids into the dilute phase. Jets
formed in this fashion expand and merge with other jets until they reach the vessel wall and are dissipated. Below the point of jet
dissipation, solids are carried upward by the jets at effective velocities which decay inversely proportional to the square root of
the distance above the bed. As the jets expand and the jet velocities decrease, some solids drop back into the dense bed.
Above the point of jet dissipation, gas velocity becomes fairly uniform and the solids entrainment becomes constant with
increasing vessel height. At this point, only those solids whose terminal velocities are equal to or less than the gas superficial
velocity continue to be entrained. The point at which the entrainment becomes constant is defined as the Transport Disengaging
Height or TDH.
For a slugging bed:
TDHS = 1.7 (UB + U − Umf ) UBS ( ft ) Eq. (13c)

TDHS = 3.1 (UB + U − Umf ) UBS (m) Eq. (13m)

where: TDHS = The transport disengaging height for a slugging bed (ft), (m)
UBS = The slug rise velocity [Eq. (7)] (ft/s), (m/s)
U = The superficial gas velocity (ft/s), (m/s)
UB = The bubble rise velocity [Eq. (6)] (ft/s), (m/s)
Umf = The superficial velocity at minimum fluidization (ft/s), (m/s)
For a bubbling bed:
TDHB = (1.5) 1.7 (UB + U − Umf ) UB ( ft ) Eq. (14c)

TDHB = (1.5) 3.1 (UB + U − Umf ) UB (m) Eq. (14m)

where: TDHB = The transport disengaging height for a bubbling bed (ft), (m)
The factor 1.5 was included in Eq. (14) based on a comparison with commercial FCC data.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 8 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

Based on very limited comparisons with measured results, Eqs. (13) and (14) predict TDH to about ± 30%.
For fluid bed systems with a non-uniform cross-section caused by diameter changes (swaged vessels) or internals such as
baffles or cyclones, the concept of a critical outage is used instead of TDH. The critical outage is defined as the vessel height
above which entrainment is constant. It can differ from the calculated TDH if changes in effective vessel cross-section occur
which affect gas velocity and entrainment.
➧ TDH, or critical outage, can be calculated with The PEGASYS cyclone module, a general purpose fluid solids program described
in Section VI-C, Appendix A. PEGASYS cyclone module first calculates bubble size and determines if bubbling or slugging
conditions control. The program then selects the proper equations to calculate TDH or the critical outage.

Entrainment Correlation
The calculation of entrainment is based on a correlation involving the square of the ratio of gas velocity to terminal velocities
(Reference 5). The correlation is:
é ù
ê ú
E n ê − 42.2 ú
= 14.5 å xi exp ê 2 ú Eq. (15)
ρg i =1 ê V − 1ú
ê U2 ú
ë ti û

where: E = The entrainment (lbs solids/ft3 gas), (kg solids/m3 gas)

ρg = The gas density (lbs/ft3), (kg/m3)
xi = The weight fraction solids in particle size increment i
n = The number of particle size increments
V = The effective gas velocity at the point at which the entrainment is calculated (ft/s), (m/s).
See further description below.
Uti = The terminal velocity for particles in increment i (ft/s), (m/s)
➧ At levels above TDH (or critical outage), V is the superficial gas velocity, U. Below TDH, V is an effective velocity, computed by
the following procedure:

1. Determine Us(X), the superficial gas velocity as a function of height X above the dense bed level. Us changes
because of changes in vessel cross-section area and presence of bed internals that also lead to an effective
decrease in available area for gas flow.
2. Determine Ve(X)
1/ 2
é TDH ù
Ve ( X ) = U ê ú ( ft / s),( m / s )
ë X û

where: Ve = The effective velocity (ft/s), (m/s)

U = The superficial gas velocity at the top of the dense bed (ft/s), (m/s)
X = The height above the dense bed (outage) (ft), (m)
TDH = Transport disengaging height (ft), (m)
3. Determine critical outage, Xc, which is the vessel height above which entrainment is constant: It is determined by
the following expression: Us(Xc)=Ve(Xc).
4. V is then given by the following expression
ìV ( X ) , X < X
V = í Ve ( X ) , X ≥ Xc ( ft/s ), (m/s) Eq. (16)
î e c c

➧ The entrainment correlation is included in PEGASYS cyclone module. For larger particles, the program uses Figure 1 to
calculate Uti with the inclusion of the Heywood shape factor z. The user can specify the shape factor, or use the program default
of 0.52 for spherical particles. For small particles, the program uses the particle clustering concept to determine Uti.

The entrainment correlation predicts measured results to ± 30%. The correlation should not be used at superficial velocities

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 9 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

➧ below 1 ft/s (0.3 m/s). For applications below 1 ft/s (0.3 m/s), consult the Process Engineering Department. A new correlation
has been developed to address entrainment at high bed superficial gas velocities, for details see document EE.701E.2002.

FLUID BED PROPERTIES, PERFORMANCE

FLUID BED DENSITY

The density of a bubbling fluidized bed can be derived from the equation for the absolute bubble rise velocity UA and the two-
phase (bubble and emulsion phases) theory of fluidization. The resulting equation is:
H ρ U − Umf
= mf = 1 + Eq. (17)
Hmf ρ UB

where H is the height of the fluidized bed and ρ its density. Subscript mf refers to minimum fluidization conditions. A more
convenient form of the equation is:
UB
ρ = ρmf (lbs / ft 3 ), (kg / m3 ) Eq. (18)
UB + U − Umf

The equation may also be used in the absence of experimental data to predict the density of a non-slugging bed. In such a case,
the bubble velocity will be given by UB = 0.7 gDB and bubble size can be estimated according to Eq. (10). Bed density at
minimum fluidization, ρmf, is usually known for the solids in question, or can be assumed to be equal to 90 to 95% of the loose
bulk density. Fluid bed density can also be expressed in terms of the bed voidage, ε, as given in Eqs. (19), (21), and (22). Non-
ideal behavior, such as gulf stream flow, will often cause actual bed density to be somewhat greater than predicted by the
equation, but no method exists to quantify the effect.

BED VOIDAGE
Another property of a fluidized bed related to density is the interstitial voidage, ε. It is related to density by the expression:
ρ
ε =1− Eq. (19)
ρp

where ρp is the particle density defined under SOLID PROPERTIES. Voidage at minimum fluidization conditions, εmf, can be
obtained from Eq. (19) using ρmf in place of ρ. In this case:
ρmf
εmf = 1 − Eq. (20)
ρp

Eq. (18) can also be used to give ρmf in terms of εmf. This gives another version of the equation for bed density:
UB
ρ = ρp (1 − εmf ) (lbs / ft 3 ), (kg / m3 ) Eq. (21)
UB + U − Umf

Voidage at minimum fluidization, εmf, is usually in the range of 0.45 to 0.5. Voidage can also be obtained from the expression:
ρ
ε =1− (1 − εmf ) Eq. (22)
ρmf

Dilute Phase Density

Dilute phase density is the concentration of solids in the dilute phase region of a fluidized bed vessel expressed as weight of
solids per volume of dilute phase. It is used to determine the holdup of solids in the dilute phase for pressure balance
calculations. Dilute phase density decreases with outage up to the critical outage. It is constant above the critical outage. Dilute
phase density is higher than entrainment because of slip, which is the tendency of the entrained solids to move through the dilute
phase region at a velocity less than the rising gas. The relationship between dilute phase density and entrainment is given by
(Reference 6):
ρ d = E (S.F.) = ρ g (lbs / ft 3 ), (kg / m3 ) Eq. (23)

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 10 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

where: ρd = The dilute phase density (lbs/ft3), kg/m3)

E = Entrainment (lbs/ft3 gas), (kg/m3 gas)
S.F. = A slip factor
ρg = The gas density (lbs/ft3), kg/m3)
The slip factor is defined as:
S.F. = Ug / US Eq. (24)

where: Ug = The gas velocity (ft/s), (m/s)

US = The solids velocity (ft/s), (m/s)
The slip factor is available from correlations of the slip factor with the ratio E / ρp, where ρp is the particle density (Reference 6).
➧ Dilute phase density is calculated by PEGASYS cyclone module.

Dilute Phase Holdup

The dilute phase holdup is the total weight of solids (lbs or kg) contained in the dilute phase and is obtained by integrating the
dilute phase density over the outage. The calculation of dilute phase density and the integration to give dilute phase holdup are
➧ made by the cyclone module in PEGASYS.. The integral equation involved in this calculation is:
xf éx ù
Holdup =
ò U
Vj ò
(E) ( A ) ê (S.F.) fd (dp )ú dX
ê ú
(lbs), (kg) Eq. (25)
o ëê o ûú

where: X = The outage (ft), (m)

Xf = The maximum outage for the calculation (ft), (m)
U = The superficial gas velocity (ft/s), (m/s)
Vj = The gas jet velocity (ft/s), (m/s)
E = The entrainment (lbs/ft3), (kg/m3)
A = The vessel area in the dilute phase region (ft2), (m2)
S.F. = The slip factor
f = A frequency function (1/ft), (1/m)
The integration is carried out in two regions, below and above the TDH (or critical outage). PEGASYS cyclone module
determines the TDH and integrates Eq. (25) accordingly.
A comparison of calculated and measured holdups based on a single test in a commercial FCC plant (Montreal) indicated the
calculated values to be about 30% low.

GAS / SOLID CONTACTING

Contacting between gas and solids in a fluidized bed reactor can affect important performance criteria such as conversion, yield
product quality, catalyst activity, temperature distribution and may possibly affect unit operability. Contacting is usually described
by the two-phase theory which states that gas moves through the bed in two phases, the bubble phase and the emulsion phase
(Reference 7).

Bubble Phase
Bubbles are gas voids relatively free of solids. Reactions which require the presence of a solid do not occur to a significant
degree in the bubbles. However, vapor phase reactions can occur in the bubbles. Since bubble velocity is related to bubble size
[Eq. (6)], the larger the bubbles the higher the upward velocity. Bubble velocities are usually greater than the overall superficial

velocity or velocity through the emulsion phase. The maximum bubble velocity is the terminal velocity of the solids. When
bubbles reach this velocity, they break into smaller bubbles with lower velocities.
The fraction of the bed occupied by bubbles, δ, can be approximated by:
U − Umf
δ≅ Eq. (26)
U − Umf + UB

Bubbles are also thought to be surrounded by a gas/solid cloud which is a region with properties intermediate between the
bubble and emulsion phases. Large, rapidly rising bubbles have thin clouds, while smaller, slower bubbles have thicker clouds.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 11 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

If the gas velocity is greater than twice the velocity at minimum fluidization, or U > 2 Umf, the bed is bubbling vigorously with
large, fast-rising bubbles. In this case, the bubble gas flow is close to plug flow.
Although bubbles are relatively free of solids, bubbles create a wake as they rise which pulls in solids, carrying them upward with
the bubbles to the top of the fluidized bed. This is the mechanism which is responsible for axial mixing of solids in a fluidized
bed. Lateral or radial mixing of solids is also caused by the bubbles. In this case, solids adjacent to the bubbles' solids are
pushed aside by the rising bubbles, then pulled into the bubble wake where they are mixed with other solids.
In summary, bubble gas flow is characterized by relatively short residence times, minimum contact with solids and is largely in
plug flow.

Emulsion Phase
The emulsion phase consists of a bubble-free, gas/solids suspension. Voidage in the emulsion phase, εe, is usually assumed to
be the voidage of minimum fluidization, εmf or about 0.5. Since solids concentration is high, reactions which require the presence
of solids occur primarily in the emulsion phase.
Solids which are carried upward in the wake of the bubbles to the top of the fluidized bed, flow downward through the emulsion
phase. The solids velocity is given by:
δ α UA
US = ( ft / s), (m / s) Eq. (27)
1− δ − α δ

where: US = The solids downward velocity (ft/s), (m/s)

δ = The bubble fraction in the bed [Eq. (26)]
α = The fraction of the bubble volume occupied by the bubble wake
UA = The absolute bubble velocity (ft/s), (m/s) [Eq. (9)]
α can be assumed to be in the range of 0.2 to 0.4.
Emulsion gas velocity is given as:
Umf
Ue = − US ( ft / s), (m / s) Eq. (28)
εmf

where: Ue = The emulsion gas velocity (ft/s), (m/s)

If bed solids are stationary, e.g., at minimum fluidization, US′ = 0 and emulsion velocity equals Umf / εmf. As solids flow is initiated
in a bubbling bed, Ue decreases and will become negative when US′ exceeds Umf / εmf. Therefore, in vigorously bubbling beds,
rapidly downward moving solids can be thought to pull emulsion gas downward, causing emulsion gas backmixing. Thus, gas
flow through the emulsion phase can range from plug flow at low gas velocities to backmixed flow at high gas velocities. This
reversal in the emulsion gas flow direction occurs when the solids downward velocity is greater than Umf / εmf or when:
U 1
> (1 − δ − α δ) (1 + ) Eq. (29)
Umf α εmf

In summary, in comparison to bubble gas flow, emulsion gas flow is characterized by longer residence times in closer contact
with the solids and can range from plug flow to backmixed flow depending on the gas velocity.

Interchange Between Bubble and Emulsion Gas

Gas flowing through a fluidized bed is not restricted to the bubble or emulsion phase, but can move between the two phases.
The interchange rate can be expressed in terms of interchange coefficients, defined as volume exchanged/volume of bed/time or
crossflow ratios, defined as the number of times bubble gas is replaced as a bubble passes through the bed. Methods to
calculate interchange coefficients and crossflow ratios are given in Reference 7.

Gas Dispersion Coefficients

The flow of gas through a fluidized bed can also be described as a diffusion or dispersion process. Dispersion coefficients
defined by standard diffusion equations have been measured and correlations developed. Equations based on the two-phase
theory have also been developed (Reference 7). The axial dispersion coefficient, Da, is dependent on α, δ, U, Umf, εmf and a
coefficient describing interchange between the bubble and emulsion phase. The radial dispersion coefficient, Dr, is dependent on
DB, and the bubble to emulsion interchange coefficient. Higher values of the dispersion coefficients, or more correctly the
dispersion numbers, indicate more complete gas mixing. The dispersion number is defined as Dε / UH where H is the fluidized
bed height. This approach combines mixing occurring in bubble and emulsion phases and mixing due to interchange between
the two phases.

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 12 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

SOLIDS MIXING AND SEGREGATION

Solids in a freely bubbling bed are usually well mixed. This results in a number of process effects. For example, well-mixed
solids result in uniform bed temperature. However, solids mixing also results in backmixing of the emulsion gas as well as the
solids and this can have a negative effect on conversion and yield. Insufficient solids mixing can cause segregation in the fluid
bed with smaller, less dense particles collecting at the top of the bed and larger, denser particles at the bottom. This can affect
entrainment rates, solids residence time, and fluidization, all of which can affect process performance.

Solids Mixing
Solids mixing in a fluidized bed is caused by bubble action as previously described. Solids are mixed axially by being pulled
upward in bubble wakes and by moving downward in the emulsion phase. Lateral or radial mixing occurs when solids displaced
laterally by a bubble are pulled into the bubble wake and mixed with other solids. Axial mixing occurs much more readily than
radial mixing. The rate solids flow across a horizontal plane in a freely bubbling bed is given by:
J ≅ α ρp (1 − εmf ) (U − Umf ) (lbs / ft 2 s), (kg / m 2 s) Eq. (30)

where: J = The solids flow rate across a horizontal plane (lbs/ft2s), (kg/m2s)
α = The fraction of the bubble volume occupied by the bubble wake (usually 0.2 to 0.4)
ρp = The particle density (lbs/ft3), (kg/m3)
εmf = The bed voidage at minimum fluidization
U = The gas superficial velocity (ft/s), (m/s)
Umf = The gas superficial velocity at minimum fluidization (ft/s), (m/s)
The degree of solids mixing can also be described by dispersion coefficients, DSA for axial and DSr for radial mixing. Larger
values of DS indicate higher degrees of mixing. DSA can be estimated from:
α 2 εmf
DSA ≅ DB (U − Umf )2 ( ft 2 / s), (m 2 / s) Eq. (31)
3 δ Umf

where: DSA = The solids axial dispersion coefficient (ft2/s), (m2/s)

DB = The bubble diameter [Eq (10)] (ft), (m)
All other terms are as previously defined
DSr can be estimated from:
3 δ Umf DB
DSr ≅ ( ft 2 / s), (m 2 / s) Eq. (32)
16 (1 − δ) εmf

The above equations apply to freely bubbling beds with U > 2Umf. Small beds operating in slug flow or beds with coarse solids
operated at velocities closer to the minimum fluidization velocity may not follow the above equations. Departures may also be
seen in very large commercial plants where mixing caused by the existence of large flow streams (gulf streaming) may occur.
The above equations should be used for general guidance. Consult the Process Engineering Department for specific cases and
for those cases where the above equations do not apply.

Gulf Streaming
In larger beds, especially if the open area of the fluidizing grid is not equally distributed, large upward flows in the central regions
of the bed can be established. The flow splits at the top, runs down the wall and recombines at the bottom. This pattern is
referred to as gulf streaming. Although gulf streaming induces effective solids mixing, it can result in very short gas residence
times which can lead to poor yields.

Segregation
Segregation is noted by distinct layers of solids of different sizes and/or densities in the fluidized bed. Segregation can occur if
the solids size distribution is bimodal, i.e., large concentrations of small and large particles with little intermediate size particles.
Appreciable differences in particle densities can also cause segregation. Larger or denser particles can accumulate on the grid,
and may defluidize, resulting in increased pressure drop, interference with the operation of internals, solids feeding and
withdrawal. Segregation in beds undergoing exothermic reactions can result in temperature runaways and bed fusion.
Segregation may also result in an accumulation of small particles at the bed surface. This can occur in beds operated close to
minimum fluidization with little bubble induced mixing.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 13 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

The tendency to segregate is dependent on the ratio of particle densities to the third power, the ratio of particle sizes to the first
power and is inversely related to U – Umf. Consult the Process Engineering Department if segregation is believed to be possible.

CHARACTERIZATION OF GAS, SOLIDS BACKMIXING

Characterizing the degree of gas and solids backmixing can be important for processes in which backmixing can have process
effects but cannot be described by either the plug flow or completely backmixed cases. This is particularly true when scaling up
data obtained in smaller pilot plants to commercial plant size. The degree of gas backmixing can be estimated using calculated
or literature values for the axial dispersion coefficients for the gas (Reference 7) or measured using gaseous tracers. The
following table gives an indication of the effect of the axial dispersion number on the degree of backmixing (Reference 8).

Daε / UH DEGREE OF BACKMIXING

0 None (plug flow)
0.002 Small
0.02 Intermediate
0.2 Large
∞ Complete

where: H = The height of the fluidized bed

All other terms are as previously defined
The degree of solids backmixing can be considered to be similar to the degree of gas backmixing, in particular if U > 2 Umf.
In addition to these effects, there are indications that backmixing increases as vessel size increases, even if the beds in the
smaller vessels are freely bubbling.
Methods are available to use this dispersion approach to determine the degree of gas and solids backmixing and the residence
time distribution of gas and solid phases in a fluidized bed.
Another method used to characterize backmixing consists of describing the fluidized bed as a series of well-mixed tanks or
stages. A single stage represents complete backmixing. A large number represents an approach to plug flow. Figure 3 shows
an arrangement which has been used to characterize gas and solid flows in a fluidized bed. Solids flows are characterized by a
series of stirred tanks with flows between tanks as shown. Gas flows are characterized by a series of stirred tanks for the
emulsion gas flow and another series for the bubble flow. Since bubble flow approaches plug flow, a larger number of tanks is
used to characterize the bubble flow. Gas can exchange between bubbles and emulsion as shown. Emulsion gas is assumed to
be in contact with the solids. The number of emulsion gas stages and solid stages are similar. The number of stages needed to
describe a particular process can be estimated based on the dispersion coefficient, can be measured by tracer tests or obtained
by back fitting experimental data.
The use of tracers is the most commonly used method to characterize the degree of gas and solids backmixing. Descriptions of
test data analysis are available (References 7 and 8). Injection of tracers into gas streams can be used to measure gas
residence times and residence time distributions. An ideal gas-phase tracer is one which can be easily detected in the outlet
stream, is completely inert at reactor conditions and is not absorbed on the solids or on the vessel walls. Solids tracers are more
difficult to use because of sampling and analysis problems. Radioactive tracers have been successfully used and should be
considered.
In addition to characterizing the degree of backmixing, it is also possible to determine the extent to which gas or solids
backmixing affects process conversions, yields and product quality. This is done by coupling the extent of backmixing to
chemical reaction rate expressions. This gives a reactor design model which can be used to analyze experimental data, scale-up
the results to commercial size equipment, and design commercial reactors.
For questions concerning the degree and impact of gas or solids backmixing on a process, consult the Process Engineering
Department.

GAS, SOLIDS FLUID DYNAMIC CALCULATIONS

Calculation methods are available which can describe flows of solids and gas in fluidized beds, transfer lines and pneumatic
conveying systems. The calculations are made with a computer based fluid dynamic flow code, FLOXC. This method is now
operational and has been used in a number of design and troubleshooting applications for FCC and Fluid Coking units. The flow
code has been used to calculate and graphically represent gas and solid flows in fluid bed risers, in regions below grids, in the
grid holes, and in the bed above grid plates, around stationary obstacles, etc.
The flow code uses physical mechanisms to describe two dimensional two phase flows, including turbulence effects, bubble flows
with bubble coalescence and breakup, and flow of solid particles. Thermal effects and chemical reactions can be included. The

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 14 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

code is particularly useful in cases where conditions occur which can cause segregated flows, impact on solid surfaces, stagnant
zones, etc.
The results of the calculations can be displayed in a number of ways, e.g., flow maps showing the flow direction and velocity for
each phase. Temperature and concentration gradients can also be displayed graphically by contour and surface plots.
Laboratory facilities are also available which can be used to simulate and visualize gas and solid flow patterns occurring in large
equipment. These facilities have been used in design improvement studies of cyclone diplegs and trickle valves, reactors, grids,
reactor termination devices, and strippers. Contact the Process Engineering Department for information on fluid dynamic
calculations or laboratory flow simulations.

SOLID PROPERTIES

PARTICLE SIZE

Size Distribution
Solid particles used in fluidized beds cover a distribution of sizes that can be characterized by statistical and graphical methods.
Particle size distributions are usually given as a cumulative distribution, i.e., particle size vs. weight, volume or number fraction of
particles smaller or larger than that particle size. Three widely used graphical representations of cumulative distributions are the
log-normal probability, normal probability and Rosin-Rammler plots.
• Log-Normal Probability - In this distribution, the logarithm of the particle sizes is normally distributed. Cumulative log-
normal distributions give a straight line on “log-probability” graph paper. The geometric standard deviation is given as:
particle size at 50% po int in distribution
s= Eq. (33)
particle size at 15.9% po int in distribution

The greater the value of s, the wider the particle size distribution. This distribution is commonly used for fluidized solids.
See Figure 4.
• Normal Probability (Gaussian Distribution) - This is the well-known distribution in which a plot of frequency vs. particle
size gives a symmetrical bell-shaped curve. Cumulative normal distribution gives straight lines when plotted on “probability”
graph paper. The mean size is the 50% point on the probability scale. The standard deviation is given as:
s = size at 50% po int − size at 15.9% po int Eq. (34)

• Rosin-Rammler Distribution - This is used to characterize solids from crushing and grinding operations. The distribution
equation is:
h
é− dù
R = 100 e ê ú Eq. (35)
ë d û

where: R = The percent solids larger than particle size d

d = A size constant called the Rosin-Rammler mean
h = A size distribution constant
For particle distributions obeying the Rosin-Rammler relation, a plot of log-log (100/R) vs. log-d results in a straight line with
a slope of h. The Rosin-Rammler mean, d , is the particle size on the plot where R = 36.8%. This representation has limited
application outside crushing and grinding.

A mean diameter commonly used in fluid solids operations is the volume surface mean particle size, also known as the
Sauter mean. It is calculated from:
1
dvs = ( ft ), (m) Eq. (36)
å xi / di

where: d vs = The volume surface mean particle size

xi = The weight fraction of the i particle size range
di = The average particle diameter in the i particle size range
Median diameters are also used in fluidization and associated gas-solids operations. These diameters, referred to as the
d50, are simply the diameter at which 50% of some property (number, weight, surface area, etc.) of the distribution is due to
particles smaller than d50. It is determined by interpolating from tabulated size distribution measurements or by reading the

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 15 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

50% point on a cumulative plot of the data. The most commonly encountered median diameters are the weight median
diameter (50% point on a weight % vs. size cumulative plot), and the number median diameter. For normal distributions, the
mean and median sizes are identical.

Measurement Techniques
There are a variety of particle size measurement techniques, each applicable over a limited range. Due to differences in
operating principles and particle properties such as shape factor, these techniques will not necessarily give identical size or size
distribution for a given sample.
• Recommendations - There is no completely satisfactory method of particle size measurement over the entire range of
particle sizes likely to be encountered. The choice depends on the size range of the particles of primary importance, the
properties and nature of the solids being measured, and how the size information obtained will ultimately be used. However,
three sizing methods, sieving, electrical conductivity and impaction are in common use within Exxon for fluid solids
operations.
Standard dry sieving is recommended for particle sizing down to the critical size for the particle matter in question. For non-
agglomerating materials such as cracking catalyst, the limit is 38 to 45 µm, the size of the smallest available screens. For
agglomerating materials such as coal char, coal ash, the lower limit by dry sieving is 75 µm. Wet sieving can be used to size
agglomerating material down to 45 µm. However, this requires more experience and equipment than dry sieving. Sonic
Sifters are recommended for small samples and to size non-agglomerating solids down to 20 µm or less depending on the
agglomerating tendency of the solids.
The other devices in general use at Exxon are the Coulter Counter and cascade impactors. Both are recommended for
those situations where they are most appropriate - Coulter Counter for laboratory sizing of smaller sized particles and
impactors for stack sampling. Comparison of size distributions obtained by each method indicates that the impactor tends to
give a wider spread in sizes, more fines due to a lower cut off size, and possibly more coarse material due to particle
agglomeration. Good agreement has been reported for spherical particles. However, the difference between results
measured by the two methods becomes greater as the particulates become less spherical. Very large differences between
size distributions measured by the two methods should not occur and if they do, this indicates the possibility of improper
sampling or sizing. Therefore, it is recommended that when cascade impactor tests are conducted, samples also be taken
for later analysis by Coulter Counter. Process Engineering Department should be consulted if conflicts between impactor
and Coulter Counter results occur. If necessary, reference methods such as image analysis or scanning electron
microscopy can be used to resolve differences.
Particle measurement methods are described in the following sections. The effective particle size range for each of the
methods described is shown in Figure 5.
Table 3 lists typical particle size distributions for a number of FCC catalysts. Other important solids properties are also
included in this tabulation.
• Sieving - Sieving is the most commonly used sizing method and can be used for solids ranging from 20 µm to 1 in. (25 mm).
In this analysis, solids are placed on top of a series of screens, each screen having smaller openings than the one above.
As the sieves are shaken, a particle falls through them until a screen is reached in which the openings are too small for the
particle to pass. Particles fine enough to pass through all of the nested screens are collected on a solid pan located under
the bottom screen.
Standard screens of either the Tyler Series of the U.S.A. sieve series are industrially used in the United States. Table 2
compares the U.S.A., Canadian, Tyler, British, French and German Standards. The designations of the U.S.A. Sieve Series
are those recommended by the International Standards Organization (I.S.O.) as an international standard. When possible,
this designation should be used when reporting a sieve analysis intended for international publication.
For a difficult-to-sieve material, e. g., one which agglomerates during dry sieving, a wet sieving approach is sometimes used
provided the material is insoluble. Material is washed through the sieves with a high pressure, low volume water stream.
Wet sieving requires that the residue on the sieves be weighed after drying, and the drying operation introduces another
source of potential error in the analysis. However, very accurate sieve analyses can be made by the wet testing method and
specially designed frames are available for this purpose.
The standard mechanically shaken dry sieving method is applicable for solids down to 45 µm (325 mesh), if the solid is free
flowing and does not agglomerate on the sieve. The wet sieving method is also applicable down to 45 µm. The ATM Sonic
Sifter uses a vertical oscillating column of air to move the solids through the screens. This system is capable of sizing free-
flowing particles down to at least 20 µm. Screens are available down to 5 µm, but the minimum measurable particle size is
usually determined by particle agglomeration on the screens.
The amount of solids used in a standard sieve analysis depends on the equipment, nature of solids and range of sizes likely
to be present in the solids. If the sample is too large, screens will be overloaded and the smaller particles may not move
down to their proper sieve. Samples too small will limit accuracy. Guidelines established by the ASTM (Reference 9)
should be used to determine the proper sample size. Sonic Sifters require a minimum sample size of 3 to 5 gm.

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 16 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

In general, the advantages of sieving are that it is a simple, relatively quick method using standardized, inexpensive
equipment. It is widely available throughout Exxon. Sonic Sifters have the added advantage of requiring a very small
sample size. A disadvantage of sieving is that the standard procedures must be followed closely to assure consistent
results. Agglomeration of the solids on the screens or attrition during shaking can also give misleading size distributions.
The cut-off size of standard sieves is also relatively high, 45 µm. Sonic Sifters are used to achieve a lower cut-off size and
wet sieving is used to prevent agglomeration. However, these methods require more experienced operating personnel and
wet sieving also requires additional time.
• Electrical Conductivity - This technique is based on the conductivity difference between particles and a suspending fluid.
The Coulter Counter from Coulter Electronics is the principal instrument used. In this instrument, particles are suspended in
an electrolyte and forced through a small aperture through which an electric current flows. Each particle displaces
electrolyte in the aperture, producing a current pulse. The amplitude of the pulse is a function of the volume of the particle
and pulse frequency is a measure of concentration. A number of electronic channels, each counting pulses of a given
amplitude, are used to determine the particle size distribution. A calibration with a monosized standard is required to set
particle sizes. The aperture sizes available allow size determinations for particles between 1 µm and 250 µm in a standard
laboratory environment. Installation is a specially controlled environment such as a clean room, permits sizing down to
0.5 µm. The Coulter Counter is commonly used for laboratory sizing of collected samples in the size range below that
measured by sieving.
The advantages of this method are that it is fast, reliable and reproducible. It requires very small sizes, less than 1 gm. It is
available in some laboratories within Exxon.
The disadvantages are that it requires skilled, experienced operating personnel, and also requires the use of multiple
apertures to cover a wide range of particle sizes. Quite often, it is combined with sieving to give a wide range of particle
sizes without using multiple apertures. It is also limited to solids which are not affected by the electrolyte used to suspend
the solids. However, most solids encountered in Exxon operations are compatible with the electrolytes now used. The
suspending electrolyte may also de-agglomerate particles which were sampled as agglomerates. This could be an
advantage or disadvantage depending on the downstream treatment the solids are subjected to.
• Impaction - Cascade impactors are devices in which a stream of suspended particles is forced to follow a series of curved
flow paths. The larger particles deviate from the flow paths due to their inertia, impact on a surface and are captured.
Smaller particles remain in the flowing stream but are also eventually captured as the stream velocity is increased and the
inertia of the smaller particles becomes sufficient to allow them to impact on a surface. Impactors usually contain seven
impaction stages on which the particles of decreasing size are captured. Cascade impactors are small, compact devices
which can be directly inserted into a duct to make insitu measurements of size distribution. They measure a size based on
the aerodynamic properties (shape and density) of the particulates over a range of 0.3 to 20 µm. Impactors are usually
backed up with a total filter to catch particles smaller than 0.3 µm and can be preceded by a small cyclone to catch particles
larger than 20 µm if they are believed to be present.
The advantages of impactors are that they are the most widely used and generally accepted method for insitu size
measurement and measure an aerodynamic size which is useful to assess performance of inertial particulate control
devices. Impactor design and use have been fairly well standardized although a number of models are available, some of
which are generally preferred over others.
A disadvantage of impactors is that they need very careful and experienced operating personnel to assure good results.
Sampling and sizing are also quite often done under less than ideal conditions which emphasizes the need for capable
personnel. As in the case of all stack sampling methods, sampling must be done under isokinetic conditions (same flow
velocity into sampler as in the duct) to give accurate results. Impactors measure agglomerated particles as the
agglomerates. This can be a disadvantage or an advantage depending on the downstream treatment the solids are
subjected to. The density of the sampled particles must also be known in order to calculate the size distribution properly.
A stack-sampling guide was assembled by Exxon Engineering (Reference 10). However, stack sampling is usually done by
contractors, and the Equipment and Environmental Engineering Department should be consulted for further information on
contractor selection.
Two other particle sizing methods are currently finding wider acceptance and are described below. These are Image
Analysis, a computer-enhanced microscopic method, and light and laser scattering and blockage.
• Image Analysis. Image analysis is a recent development which couples with a light or electron microscope to give
measurements not only of particle size, but also shape, morphology, texture, color, etc. The computer performs the time-
consuming and tedious tasks which formerly limited the use of the microscope for particle sizing and characterization. The
image analyzer automatically scans a projected image of the particles, or a microscopic photograph of an array of particles,
and calculates particle size distributions, shape factors, length-to-diameter ratios, roughness, sphericity, etc. It can
differentiate between particles of different shapes and count only those particles with a specified shape, e.g., excluding filter
fibers from a sample of cracking catalyst. The range of particle sizes measured by this method is that of the microscope
system itself. For a light microscope, this range is 0.5 to 300 µm. For an electron microscope, sizes down to 0.001 µm can
be measured.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 17 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

The advantages of this system are that it provides all the additional information offered by microscopic characterization
quickly and at low cost. Further, it enables calculation of other properties such as roughness, length-to-diameter ratios, etc.,
which cannot be done manually in standard microscopy. It also requires a very small sample size.
The disadvantages are that it looks only at projected areas of the particles which could be the maximum area if the particles
are not isotropic, and requires skilled operators to prepare samples properly and operate the system. Image analyzers are
also costly and not in widespread use.
• Light Scattering and Blockage. In techniques based on light scattering, dispersed particulates are passed through a light
or laser beam and the light scattered by the particles in the stream is detected by a photomultiplier. With a gaseous medium,
particle sizing is possible for a range of sizes between 0.3 and 50 µm and with liquids a size range of about 0.5 to 600 µm is
measurable. Some of the more popular instruments encountered in fluidization based on this principle are the Malvern
particle and droplet sizer, the Leeds and Northrup Micotrac, the HIAC automatic particle counter, the Royco particle counter,
Bausch and Lomb 40 - 1, the Climet Particle Analyzer, and the Cilas Granulometer. Both laboratory and on-line devices are
available.
For laboratory applications these instruments offer the advantages of rapid and reliable measurement with little operator
training required. No special liquid is required for operation with liquids, and calibration is not necessary once the light or
laser beams have been aligned with the measurement cell and detector. The Microtrac, Malvern, and Granulometer are
based on analysis of Fraunhofer diffraction patterns generated from the light scattering. Each features a personal computer
with software to generate particle size distributions, display results graphically, and calculate various characteristic sizes for
the distribution.
An advantage of an on-line model is that it is a non-intrusive device. Light or laser beams are focused on the moving
gas/solids stream through windows.
A disadvantage of these systems is that they can analyze only for solids in a given range and need multiple measurements
to cover the widest range. This is a particular disadvantage in an on-line system since the device must look at a number of
different samples to cover a wider range. The laboratory models require redispersion of the solids which raises questions
similar to those encountered with the Coulter Counter. If samples contain agglomerates, the on-line system detects them as
agglomerates but the laboratory system may de-agglomerate the particles and count them as primary particles. Again, this
could be an advantage or disadvantage depending on downstream treatment of the particles. This system also sees only a
projected area as does microscopy. Light and laser scattering systems are still costly and are not in widespread use as yet,
although a few systems are in operation in Exxon.
Three other methods used to size particles are centrifugation, elutriation, and sedimentation. These have been used
extensively in the past but are no longer in widespread use, especially for gas fluidized solids. They are described below.
• Centrifugation - Centrifugal classification can be used to determine powder particle size distributions. Particle settling rates
are accelerated by centrifugal forces which cause particulates to deposit at positions in the classifier which depend on the
settling velocity of the particle. This device measures aerodynamic sizes and is generally used for powders having
aerodynamic diameters between 2 and 50 µm. The most widely used commercial centrifugal particle separator is the Bahco
particle classifier. This device is very time-consuming and each device must also be calibrated. Measured size distributions
are affected by particle agglomeration, electrostatic effects and other surface properties. It is not widely used in Exxon.
• Elutriation - Devices based on the elutriation principle are used for size analysis of dry particulates. Particles are introduced
into an upflowing fluid stream, and the particles with a settling velocity less than the fluid velocity are carried upward or
elutriated by the fluid. By subjecting the sample to a series of increasing air velocities, a particle size distribution is obtained.
The Roller classifier is an example of this type device. These devices result in an aerodynamic size measurement and are
primarily used for particulates with sizes ranging from 5 to 200 µm.
Results are affected by particle properties, time for analysis is long and operation requires experienced personnel. In the
1950's, the Roller classifier was the primary method for analysis of FCC catalyst, but it is no longer in widespread use.
• Sedimentation - In sedimentation devices, particles are allowed to settle at the rate or velocity determined by their effective
size and density. The sizes measured by these devices are effective sedimentation or settling diameters the same as those
in elutriation instruments and the size range is usually about 1 to 500 µm in liquids and about 2 to 200 µm in gas
sedimentation devices. The most widely used gas sedimentation device is the Sharples Corporation Micromerograph.
The disadvantages of this system are similar to those of elutriation devices. In addition, the devices are very large and do
not produce very reproducible results.
Liquid/solid sedimentation devices such as the Sedigraph are more widely used to characterize solids found in liquid/solid
systems. This device is more reproducible than gas/solid sedimentation devices.
• Measurement Capabilities - A particle characterization laboratory equipped to measure particle size, density, pore size
distribution, morphology, and composition in a clean room environment is operated in the Process Engineering Department.
Equipment available for particle characterization is listed below.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 18 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

PORE SIZE PARTICLE MORPHOLOGY,

PARTICLE SIZING PARTICLE DENSITY DISTRIBUTION COMPOSITION
Sieves Mercury Porosimeter Mercury Porosimeter Image Analyzer
Sonic Sifter Liquid Pycnometer SEM/EDAX
Coulter Counter Helium Pycnometer Optical Microscope with
Sedigraph Hot Stage, Video
Malvern Particle, Recording
Droplet Analyzer
Image Analyzer
Optical Microscope
Scanning Electron
Microscope

SOLID DENSITY
Important solid density properties include skeletal density, particle density, bulk density and two related properties, pore volume
and bed void fraction.

Skeletal Density
The skeletal density, ρs, is the density of a single particle excluding pores, that is the density of the “skeleton” of the particle if the
particle is porous. It is sometimes referred to as the “true” density. For non-porous particles, it is equivalent to the particle
density, and for porous particles, it is higher than the particle density.
Skeletal density can be measured by liquid or gas pycnometers. With liquids, a pycnometer with a fixed and known volume is
used. A measured weight of solids is immersed in a liquid of known density which wets the solid and penetrates into the pores of
the particles. The volume of liquid displaced by the solids is then determined by difference. The skeletal density is the weight of
solids divided by the displaced volume. Water is most commonly used in liquid pycnometer work. Where the solids are not
completely wetted by water or float in water (e.g., coke and coal), Varsol can be used.
Gas pycnometers are less commonly used because they are more elaborate and expensive, but may give more accurate results
due to better penetration of the pores of the solids by the gas.

Particle Density
This is the density of a particle including the pores or voids within the particles. It is the weight of the particle divided by the
volume occupied by the entire particle as if an invisible membrane were surrounding the particle. The particle density is
sometimes referred to as the “apparent” density and is related to the skeletal density:
1 + ρf (P.V.)
ρp = (lbs / ft 3 ), (kg / m3 ) Eq. (37)
1 + (P.V.)
ρs

where: ρp = The particle density (mass/vol)

ρs = The skeletal density (mass/vol)
P.V. = The pore volume per unit mass of solids (vol/mass)
ρf = The density of the fluid contained in the pores of the solids (mass/vol)
➧ If the particle pores are filled with a gas, then ρf (P.V.) becomes negligibly small:
1 1 Eq. (38)
= + P.V. (ft3 /lb), (m3 /kg)
ρp ρs

For a porous particle, the particle or apparent density is lower than the skeletal density. For non-porous particles, the pore
volume is zero and the skeletal and particle densities are identical.
Particle density measurement can be made directly by immersion in a non-wetting fluid such as mercury which does not
penetrate into the pores. This technique is similar to the liquid pycnometer methods used for skeletal density determination
except that the volume of fluid displaced is that for the entire particle and not just the skeleton.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 19 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

Bulk Density
Bulk density is the overall density of a bed of particles, including the interparticle voids. It is most easily measured by pouring a
weighed sample of particles through a funnel into a graduated cylinder. The volume occupied determines the loose bulk density.
If the graduated cylinder is then gently tapped or vibrated, the volume decreases as the particles become more closely packed.
The minimum volume from tapping determines the tapped bulk density.
Bulk density is related to particle density by the relation:
ρB = ρp (1 − ε) (lbs / ft 3 ), (kg / m3 ) Eq. (39)

where ρB is bulk density and ε is the interparticle void fraction in the sample. Although wide variations of ε are possible,
depending on the nature of the particles, most particles of interest in fluidization have a loosely poured voidage of about 0.4 to
0.45. Tapped voidages are about 10% lower.
Of greatest interest in fluidization application is the bulk density at minimum fluidization conditions. This is usually 90 to 95% of
loose bulk density. It is measured in a fluidization column as the bed weight divided by its volume when the gas velocity is just
high enough to fluidize the particles.
The following table summarizes some useful density ratios.

DENSITY RATIO VALUE

ρmf / ρBL 0.9 - 0.95
ρp / ρBL 1.7 - 1.8
ρp / ρBT 1.4 - 1.6
ρmf / ρp 0.5 - 0.6

where: ρmf = The density at minimum fluidization

ρBL = The loose bulk density
ρBL = The tapped bulk density

Pore Volume
Pore volume can be determined from the adsorption and/or the desorption isotherms of equilibrium quantities of gas absorbed or
desorbed at different relative pressures. Mercury intrusion is another approach to obtaining the pore volume. In this approach,
hydrostatic pressure is used to force mercury into the pores to generate a plot of penetration volume versus pressure. The size
of the pore openings is related to the pressure, and this method gives pore size distribution as well as total pore volume. Both
physical adsorption and mercury intrusion require expensive equipment and a trained operator. For cat cracking catalyst, a
simple technique has been developed based on the fact that cat cracking catalyst is normally very freely flowing, absorbs water
preferentially into the pores of the catalyst, and becomes cake-like and is not free flowing when water is trapped between catalyst
particles. In this technique, a weighed amount of catalyst is placed on the bottom of a float-bottomed jar. Water is added
dropwise, the catalyst and water are shaken in the capped jar, and the bottle is turned upside down periodically. While the water
is penetrating the pores, the catalyst remains free flowing. Once the pores become filled, the water begins to fill the voids
between particles, the catalyst cakes up, and remains on the bottom of jar when the jar is inverted. The pore volume is thus
equal to the volume of water added just before the catalyst begins to cake divided by the weight of catalyst in the jar. Knowing
the pore volume, the particle density can be calculated (in gas) by using Eq. (38).

SOLIDS FLOW PROPERTIES

Solids flowability properties include the angle of repose, angle of internal friction, and the angle of slide. These parameters are
used to classify solids conveyability and ability to flow from hoppers and bins and to design equipment for solids flow and storage
vessels.

Angle of Repose
This is the angle between a horizontal plane and top of a pile of solids. The poured angle of repose is obtained when a pile of
solids is formed, while the drained angle results when solids are drained from a bin. Figure 6 shows both versions of the angle of
repose. For solids with a uniform particle size, the drained and poured angles are about the same; however, when the solids
have a wide particle size distribution, the drained angle is higher than the poured angle. The drained angle is more significant in
bin design. The differences between the angles are rarely important and since the poured angle is the easier angle to measure,
it is the most frequently reported. The lower the angle of repose the more free flowing the material and the shallower the bin
angle required. Very free-flowing granules are those having an angle of repose varying from 25° to 30°; free-flowing granules

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 20 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

have an angle between 30° and 38°; fair or passable flow for powders have an angle of 38° to 45°; finer, cohesive powders have
an angle between 45° and 55°; and very cohesive powders have angles from 55° to 70°. Moisture increases the angle of repose.
The variation of angle of repose with moisture content is likely due to the surface layer of moisture that surrounds each particle
and surface tension effects which hold aggregates of solids together.

Angle of Internal Friction

The angle of internal friction is the equilibrium angle between the moving particles of a material and bulk or stationary solids in a
hopper or bin as shown by the angle α in Figure 7. The angle of internal friction is invariably greater than the angle of repose.
Although it is the more significant design parameter, it is infrequently reported due to measurement difficulties.

Angle of Slide
This is the angle from the horizontal of an inclined surface on which an amount of material will slide downward due to its own
weight. The angle of slide is sometimes used in chute and hopper design, but it also has implications in the pick-up or movement
of solids in a gas conveying system. It is a measure of the relative adhesiveness of a dry material to a dissimilar surface.
The angle of slide depends on the nature of the solids, type of surface, surface condition (roughness, cleanliness, dryness),
surface configuration (degree of curvature), the way the material is placed on the surface, and the rate of change of the slope of
the surface during testing.

ATTRITABILITY
Attritability of a solid is a measure of its tendency to break or to be ground during processing. Fine particles formed by attrition
can affect fluidization properties, increase entrainment and solids loss rates. In a fluidized bed in which the solids do not undergo
chemical reactions, attrition occurs primarily at the grid due to the effect of high velocity jets. Less attrition occurs due to particle-
wall and particle-particle impact. Significant attrition may also occur in cyclones due to high velocity particle-wall collisions. In
processes where solids are undergoing reactions such as gasification or combustion, decrepitation can be caused by the reaction
itself. Attrition can also occur in feeding and conveying systems and possibly in slide valves.
Attritability due to high velocity jets is usually measured in a standard test by exposing the particles to the action of a gas jet in a
small fluidized bed. The production of fines, i.e., particles less than a predetermined particle size, is measured by retaining all the
solids and measuring size distribution before and after the test. Facilities for standard jet attrition tests are available in the
Process Engineering Department.
Attrition caused by passage of the solids through a cyclone can be measured by cycling the solids through a cyclone a number of
times and measuring the production of fines. Systems for measuring impact attrition in cyclones and in other impact zones are
also available in the Process Engineering Department. Samples of at least 5 gal. are required.
Jet attrition rates are correlated using an attrition horsepower concept. Attrition, which is measured by the production of fines
smaller than a certain size, e.g., 20 µm, can be correlated with the product of particle kinetic energy and flow rate, or horsepower.
Attrition rate = f ( N, ρg, U3 , do2 ) Eq. (40)

where: N = The number of grid holes

ρg = The gas density
U = The gas velocity through the grid holes
do = The diameter of the grid holes
Impact attrition, measured as the percentage fines produced per pass through the attrition test unit, is correlated with U3. The
impact attrition rate is then related to the product of the circulation rate and percentage fines produced per pass. At higher solids
loading, the measured rate decreased due to a solids cushioning effect and the correlation must be modified. Separate
measurements of jet and impact attrition must be made since a relationship between the two types of attrition has not been
developed. Attrition tests are used to measure the distribution of fines produced in the test as well as the gross attrition rate.
Fines distribution patterns differ with different solids. For example, some solids produce a bimodal distribution of fines with a
relatively higher concentration in the finest particle size range. Serious errors in entrainment calculations can occur if the proper
fines distribution is not used.
The effect of attrition on particle size distribution, entrainment and solids losses in a fluidized bed process can be calculated using
a population balance computer model, DYNAMOD, available in the Process Engineering Department. It is described in Section
VI-C, Particle Collection. A newer version of this program Particle Balance Model (PBM) is also available.

EROSION
Equipment used in fluid solids processes can be damaged by erosion resulting from contact with moving solids. The amount of
erosive wear is due in part to the properties of the moving solids, but is also affected by the nature of the target (wearing) material
and conditions occurring at the point where the moving solids impact the wearing surface.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 21 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

Types of Erosion
Materials undergo either ductile or brittle erosion. Ductile erosion rates go through a maximum at lower impingement angles by
the erodents, usually 20° to 30° from the horizontal. Wear is usually by plowing, cutting and gouging of the surface. As the name
implies, ductile materials such as metals are susceptible to this type of erosion. Brittle erosion rates are at a maximum at a 90°
impingement angle. Wear is usually by cracking and fracturing of the surface. Brittle materials such as refractories and ceramics
are susceptible to this type of erosion. Figure 8 illustrates the difference between ductile and brittle erosion by showing the
response of a ductile material (aluminum) and a brittle material (aluminum oxide) to the impingement angle.

Factors Affecting Erosion

Erosion rates are affected by the following factors:
1. Impingement Angle - This was described above. Some materials will exhibit some degree of both ductile and brittle erosion
properties.
2. Particle Velocity - Wear is proportional to the kinetic energy of the particle, i.e., to 1/2 mv2. This relationship has been
confirmed in a number of experimental studies which have measured values of the velocity exponent close to 2. However,
some systems have given exponents ranging up to 4, especially if the target is a refractory or ceramic material.
3. Particle Size - Wear increases with particle diameter above a threshold diameter of 10 to 20 µm, up to a limiting diameter of
100 to 120 µm, where wear then becomes constant. The decrease in wear rate above the limiting diameter is possibly due
to particle fragmentation effects which occur at larger particle sizes. Wear is proportional to the kinetic energy of the
particles, indication that wear should be proportional to particle diameter cubed. Correlation of erosion rates on refractory
material with particle diameter gave exponents of 1.6 to 2.5 (References 11 and 12). However, this exponential relationship
can only be used over the range of test data because of the threshold and limiting diameter effects. There is an indication
that erosion in bends may be greater with smaller particles.
4. Particle Concentration - The effect of particle concentration is not clear. Wear rates increase in some cases with
concentration and decrease in others. High concentrations are believed to result in a protective barrier around the target
caused by rebounding particles which can then decrease erosion rates. However, the measured effect of particle
concentration may be confounded by particle acceleration effects (Reference 12). Particle velocities at higher
concentrations may be lower than at lower concentrations due to particle inertia. Lower particle velocity will result in less
erosion.
5. Particle Hardness and Shape - Wear increases with particle hardness. Angular particles also cause increased wear, but
angular particles are also harder, so the effect of angularity and hardness are interrelated. Sharp particles do increase brittle
erosion rates. In ductile erosion, rounded particles cause wear by plowing, while angular particles cause wear by cutting.
Cutting wear is usually more severe than plowing. Erosion rates measured with a number of cracking catalysts varied by a
factor of more than 2, depending on the catalyst type and history (Reference 11).
6. Particle Rotation - The effect of rotation depends on particle shape. Rotation of angular particles increases wear.
7. Nature of the Carrier Fluid and Temperature - Corrosive environments can result in very high erosion rates due to a
synergistic erosion/corrosion effect. Erosion/corrosion can occur in FCC reactors due to sulfur induced corrosion. Erosion
generally increases with temperature, apart from the possible effect of temperature on corrosion rates. In some cases,
erosion rates have decreased with increasing temperature. Increasing pressure in a corrosive coal gasification reactor
environment resulted in increased corrosion/erosion by char particles. Erosion in a slurry is a complicated matter, affected
by a number of variables.
8. Target Surface Properties, Shape, Stress Level - These effects are not well understood. Erosion usually decreases with
target hardness except where the increased hardness is obtained by work hardening of the surface. In this case, there is
little effect on erosion. Wear usually decreases as the melting point of the metal and elastic modulus of the metal increases.
Dense refractories are more resistant to brittle erosion since density increases hardness. For heterogeneous materials
(most refractories), the less resistant phases are preferentially eroded, undermining the more resistant phases which can
then fall out.

Current State of Technology, Recommendations

There is currently a lack of understanding of the fundamentals of erosion. Laboratory measurements of erosion have not been
standardized and test data cannot always be properly interpreted. Most useful information is empirical, usually based on
operating experience. Consult Sections VI-A and VI-C, and International Practices IP 19-3-2 for further information. The
understanding of the factors affecting erosion, described above, can be used to guide recommendations for new applications.
Tests such as ASTM C704 can also be used to give comparative rating of materials. Consult the Equipment and Environmental
Engineering Department for design guidelines.

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 22 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

SAMPLE CALCULATIONS

MINIMUM FLUIDIZATION VELOCITY, VOLUME, SURFACE MEAN PARTICLE DIAMETER

Given: Particle size distribution in Figure 4
ρs = 165 lbs/ft3
µs = 6.72 x 10-6 lb/ft/s
ρs = 0.44 lbs/ft3
U = 3 ft/s
• Calculate dp as dvs, using Eq. (36)

Wt. frctn xi µm)

di (µ xi / di
0 - 10 0.1 28 0.0036
10 - 20 0.1 36 0.0028
20 - 30 0.1 43 0.0023
30 - 40 0.1 48 0.0021
40 - 50 0.1 53 0.0019
50 - 60 0.1 58 0.0017
60 - 70 0.1 65 0.0015
70 - 80 0.1 74 0.0014
80 - 90 0.1 83 0.0012
90 - 100 0.1 110 0.0009
Σ xi / di = 0.0194

1
dp = dvs = = 52 µm = 1.71 x 10 − 4 ft
å i / di
x

• Calculate Umf using Eq. (1)

1/ 2
Umf dp ρ g é 0.0408 dp3 ρ g (ρ s − ρ g ) g ù
= ê(33.7 )2 + ú − 33.7
µg ê µ 2g ú
ë û

1/ 2
é 0.0408 (1.71 x 10 − 4 )3 x 0.44 x (165 − 0.44) x 32 ù
= (33.7)2 + ê ú − 33.7
êë (6.72 x 10 − 6 )2 úû

= 0.155

0.155 x 6.72 x 10 −6
Umf = = 0.014 ft / s
1.71 x 10 − 4 x 0.44

dp Umf ρ g 1.71 x 10 −4 x 0.014 x 0.44

Rep = = = 0 .2
µg 6.72 x 10 − 6

Eq. (2) is applicable since Rep < 20.

Check using Eq. (2).
dp2 (ρ s − ρ g ) g
Umf =
1650 µ g

(1.71 x 10 −4 )2 (165 − 0.44 ) x 32

=
1650 x 6.72 x 10 − 6

= 0.014 ft/s

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 23 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

TERMINAL VELOCITY
Given: ρp = 101 lb/ft3
Obtain dp as d p50 from Figure 4
dp50 = 56 mm = 1.84 x 10-4 ft
• Calculate single particle terminal velocity, Uto, using Eq. (4). Check Rep to see if Eq. (4) is applicable.
dp2 (ρp − ρ g ) g
Uto =
18 µ g

(1.84 x 10 −4 )2 (101 − 0.44 ) x 32

=
18 x 6.72 x 10 − 6

= 0.90 ft/s

dp Uto ρ g
Rep =
µg

1.84 x 10 −4 0.8 x 0.44

=
6.72 x 10 − 6

= 11
Since Eq. (4) applies for Rep less than 1, use Figure 1 to calculate Uto.
dp
X = 1/ 3
é 3 µ2g ù
ê ú
ê 4 ρp (ρp − ρg ) g ú
ë û

1.84 x 10 −4
X = 1/ 3
é 3 (6.72 x 10 − 6 )2 ù
ê ú
ëê 4 x 0.44 x (101 − 0.44 ) x 32 úû

= 6.3

Uto
Y (Figure 1) = 0 .8 = 1/ 3
é 4 µ g (ρp − ρ g ) g ù
ê ú
ê
ë 3 ρ 2g ú
û

Uto
Y = 1/ 3
−6
é 4 x 6.72 x 10 (101 x − 0.44 ) 32 ù
ê ú
êë 3 (0.44 )2 úû

Uto
= = 0. 8
0.530

Uto = 0.42 ft/s

• Calculate Uet, effective terminal velocity based on clustered particles. Obtain cluster index n, from Figure 2.
n = 7
X (Figure 1) = 6.3 x 7 = 44

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 24 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

Y (Figure 1) = 8
Uet = 8 x 0.53 = 4.2 ft/s

BUBBLE DIAMETER, RISE VELOCITY

• Calculate bubble diameter, DB, using Eq. (10). First, calculate Recl [Eq. (11)] to see if Eq. (10) is applicable.
dcl Utcl ρ g
Recl =
µg

dcl = 0.264 (dp )0.73

= 0.264 (1.84 x 10 −4 )0.73

= 5.0 x 10 −4 ft

5 x 10 −4
n = = 2 .7
1.84 x 10 − 4

Utcl = 1.6 ft/s (Figure 1)

5 x 10 −4 x 1.6 x 0.44
Recl = = 52
6.72 x 10 − 6

Eq. (10) is applicable since 0.4 < Recl < 500

0.666
é (ρp − ρ g )2 ù
DB = 0.03 ê ú ( dp )1.46
ê ρg µ g ú
ë û

0.666
é (101 − 0.44)2 ù
= 0.03 ê ú (1.71 x 10 − 4 )1.46
−6
ëê 0.44 x 6.72 x 10 ûú

= 0.21 ft
• Calculate bubble rise velocity, UB, using Eq. (6)
UB = 0.71 g DB

= 0.71 32 x 0.21

= 1.8 ft/s

TRANSPORT DISENGAGING HEIGHT

Given: U = 3 ft/s
• Calculate TDH for a bubbling bed using Eq. (14)
TDH = (1.5) 1.7 (UB + U − Umf ) UB

= (1.5) 1.7 (1.8 + 3.0 − 0.014) 1.8

= 16 ft

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 25 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

DENSITY, VOIDAGE
Given: ρmf = 56 lbs/ft3
ρρ = 101 lbs/ft3
U = 3 ft/s
• Calculate the density, ρ, of a bubbling bed using Eq. (18)
UB
ρ = ρmf
UB + U − Umf

1 .8
= 56
1.8 + 3.0 − 0.014

= 21 lbs/ft3
• Calculate bed voidage, ε, using Eq. (19)
ρ
ε = 1−
ρp

21
= 1−
101

= 0.79

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 26 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

NOMENCLATURE
A = Vessel area in dilute phase region, ft2 (m2)
D = Bed diameter, ft (m)
Da = Axial dispersion coefficient, ft2/s (m2/s)
DB = Bubble diameter, ft (m)
Dr = Radial dispersion coefficient, ft2/s (m2/s)
DSA = Solids axial dispersion coefficient, ft2/s (m2/s)
DSr = Solids radial dispersion coefficient, ft2/s (m2/s)
d = Rosen-Rammler mean particle size constant
di = Average particle diameter in size increment i, ft (m)
dp = Particle diameter, weight mean or volume surface, ft (m)
dvs = Volume surface mean particle size, ft (m)
d50 = Particle diameter of solids at which 50% of some property is due to particles smaller than d50
E = Entrainment, lbs solids/ft3 (kg solids/m3)
f = Frequency function, 1/ft (1/m)
➧ g = Gravitational acceleration, 32.2 ft/s2 (9.81 m/s2)
H = Fluidized bed height, ft (m)
Hmf = Fluidized bed height at minimum fluidization, ft (m)
h = Height above grid, ft (m)
h = Rosen-Rammler size distribution constant Eq. (35)
J = Solids flow rate across horizontal plane, lbs/ft2s (kg/m2s)
n = Number of particle size increments or cluster index
P.V. = Pore volume per unit mass of solids, vol/mass
R = Percent solids larger than particle size d
Recl = Reynolds number based on particle cluster diameter
Rep = Reynolds number based on particle diameter
s = Standard deviation
S.F. = Slip factor
TDH = Transport disengaging height, ft (m)
TDHB = Transport disengaging height for a bubbling bed, ft (m)
TDHS = Transport disengaging height for a slugging bed, ft (m)
U = Superficial gas velocity, ft/s (m/s)
UA = Bubble rise velocity relative to stationary observer, ft/s (m/s)
UB = Single bubble rise velocity relative to solids velocity, ft/s (m/s)
UBS = Slug rise velocity relative to solids velocity, ft/s (m/s)
Ue = Emulsion gas velocity, ft/s (m/s)
Umf = Superficial gas velocity at minimum fluidization ft/s (m/s)
US = Solids velocity, ft/s (m/s)
US′ = Solids downward velocity, ft/s (m/s)
Ui = Gas jet velocity, ft/s (m/s)
Ut = Solids terminal velocity, ft/s (m/s)
Utcl = Terminal velocity of particle cluster, ft/s (m/s)
Uti = Terminal velocity of particles in size increment i, ft/s (m/s)
Uto = Terminal velocity of single particle, ft/s (m/s)

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 27 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

V = Effective gas velocity where entrainment is calculated, ft/s (m/s)

X = Height above dense bed (outage), ft (m)
xi = Weight fraction solids in particle size increment i
Z = Heywood shape factor
α = Fraction of bubble volume occupied by bubble wake
α = Angle of internal friction, angle of impingement
β = Angle of repose
∆P = Pressure differential, lbs/in.2 (kPa)
δ = Fraction of bed volume occupied by bubbles
ε = Bed interstitial voidage fraction
εe = Emulsion phase voidage fraction
εmf = Bed voidage fraction at minimum fluidization
µ = Gas viscosity, lbs/ft/s (kg/m/s)
ρ = Fluidized bed density lbs/ft3 (kg/m3)
ρB = Bulk density of solids lbs/ft3 (kg/m3)
ρd = Dilute phase density, lbs/ft3 (kg/m3)
ρf = Density of fluid in solids pores lbs/ft3 (kg/m3)
ρmf = Fluidized bed density at minimum fluidization, lbs/ft3 (kg/m3)
ρp = Particle density, lbs/ft3 (kg/m3)
ρs = Solid skeletal density, lbs/ft3 (kg/m3)

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 28 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

TABLE 1
HEYWOOD SHAPE FACTORS FOR REPRESENTATIVE SOLIDS

SOLID SHAPE FACTOR, Z

Spheres 0.52(1)

FCC Catalyst 0.52

Sand 0.25

Coal 0.23

Oil Shale 0.15 - 0.50

Limestone 0.16

Flake Graphite 0.023

Mica 0.003

Note:
π
(1) Shape factor for sphere = = 0.52
6

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 29 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

TABLE 2
COMPARISON OF VARIOUS STANDARD SIEVE SERIES

U.S.A.(1), CANADIAN(2) TYLER(3) BRITISH(4) FRENCH(5) GERMAN(6)

MESH NOMINAL NOMINAL

*STANDARD ALTERNATE DESIGNATION OPENING MESH NO. OPENING NO. OPENING
125 mm 5 in.
106 mm 4.24 in.
100 mm 4 in.
90 mm 3 1/2 in.
75 mm 3 in.
63 mm 2 1/2 in.
53 mm 2.12 in.
50 mm 2 in.
45 mm 1 3/4 in.
37.5 mm 1 1/2 in.
31.5 mm 1 1/4 in.
26.5 mm 1.06 in. 1.05 in.
25.0 mm 1 in. 25.0 mm
22.4 mm 7/8 in. 0.883 in.
19.0 mm 3/4 in. 0.742 in. 20.0 mm
18.0 mm
16.0 mm 5/8 in. 0.624 in. 16.0 mm
13.2 mm 0.530 in. 0.525 in.
12.5 mm 1/2 in. 12.5 mm
11.2 mm 7/16 in. 0.441 in.
10.0 mm
9.5 mm 3/8 in. 0.371 in.
8.0 mm 5/16 in. No.2 1/2 8.0 mm
6.7 mm 0.265 in. 3
6.3 mm 1/4 in. 6.3 mm
5.6 mm No.3 1/2 3 1/2
5.00 mm 38 5.00 mm
4.75 mm 4 4
4.00 mm 5 5 4 00 mm 37 4.00 mm
3.35 mm 6 6 3.35 mm 5
3.15 mm 36 3.15 mm
2.80 mm 7 7 2.80 mm 6
2.36 mm 8 8 2.40 mm 7 2.50 mm 35 2.50 mm
2.00 mm 10 9 2.00 mm 8 2.00 mm 34 2.00 mm
1.70 mm 12 10 1.68 mm 10 1.60 mm 33 1.60 mm
1.40 mm 14 12 1.40 mm 12
1.25 mm 32 1.25 mm
1.18 mm 16 14 1.20 mm 14
1.00 mm 18 16 1.00 mm 16 1.00 mm 31 1.00 mm
850 µm 20 20 850 µm 18
800 µm 30 800 µm
710 µm 25 24 710 µm 22
630 µm 29 630 µm
600 µm 30 28 600 µm 25
500 µm 35 32 500 µm 30 500 µm 28 500 µm
425 µm 40 35 420 µm 36
400 µm 27 400 µm
355 µm 45 42 355 µm 44
315 µm 26 315 µm
300 µm 50 48 300 µm 52

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 30 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

TABLE 2 (Cont)
COMPARISON OF VARIOUS STANDARD SIEVE SERIES

U.S.A.(1), CANADIAN(2) TYLER(3) BRITISH(4) FRENCH(5) GERMAN(6)

MESH NOMINAL NOMINAL

*STANDARD ALTERNATE DESIGNATION OPENING MESH NO. OPENING NO. OPENING
250 µm 60 60 250 µm 60 250 µm 25 250 µm
212 µm 70 65 210 µm 72
200 µm 24 200 µm
180 µm 80 80 180 µm 85
160 µm 23 160 µm
150 µm 100 100 150 µm 100
125 µm 120 115 125 µm 120 125 µm 22 125 µm
106 µm 140 150 105 µm 150
100 µm 21 100 µm
90 µm 170 170 90 µm 170
80 µm 20 80 µm
75 µm 200 200 75 µm 200 71 µm

63 µm 230 250 63 µm 240 63 µm 19 63 µm

56 µm
53 µm 270 270 53 µm 300
50 µm 18 50 µm
45 µm 325 325 45 µm 350 45 µm
40 µm 17 40 µm
38 µm 400 400

Notes:
(1) U.S.A. Sieve Series - ASTM Specification E-11-70.
(2) Canadian Standard Sieve Series 8-Gp-Id.
(3) Tyler Standard Screen Scale Sieve Series.
(4) British Standards Institution, London BS-410-62.
(5) French Standard Specifications, AFNOR X-11-501.
(6) German Standard Specification DIN 4188.
* These sieves correspond to those recommended by ISO (International Standards Organization) as an International Standard and this
designation should be used when reporting sieve analysis intended for international publication.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 31 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

TABLE 3
REPRESENTATIVE FLUIDIZED SOLIDS PROPERTIES
FCC CATALYST SIZE DISTRIBUTION
(Measurements by Coulter Counter except as noted)

VENDOR DESIGNATION SOURCE DATE SIZE (MICRONS) AT WT% LESS THAN

20% 50% 80%
Davison CBZ-1 - equil Bayway 9/84 42 60 79
CCZ-200 - fresh CRA 1/82 40 65 95
DA-250 - equil Davison 6/83 48 70 100
DA-400 - fresh BR-PCLA-3 10/86 38 54 75
DA-400 - equil Baton Rouge 6/84 36 52 74
DA-400 - equil Baton Rouge 9/84 43 60 85
DA-400 - equil BR-PCLA-3 5/86 39 53 73
GXO-25 + - fresh Davison 6/85 42 63 89
GXO-25 + - equil Bayway 10/85 40 54 80
RC-22 - fresh Southwestern 12/81 42 69 110
Super DX - fresh Lake Charles 4/83 44 65 89
Super DX - fresh Valley Field 4/83 47 70 100
Super DX - fresh Campana 11/81 38 70 113
Super DX - equil Benicia 9/84 44 61 85
Super Nova DE - fresh BR-PCLA-2 10/86 37 53 78
Super Nova DE - fresh Davison 1/87 43 62 88
Super Nova DE - equil BR-PCLA-2 8/86 37 51 69
3B-4 - equil Port Jerome 9/74 41 52 68
Engelhard HFZ-20 - equil Engelhard 1/85 43 57 77
HFZ-33 - fresh Engelhard 1/85 39 58 82
HFZ-33 - equil Engelhard 1/85 43 59 82
UDP-1560 - fresh Englehard 11/86 48 68 92
UDP - fresh Englehard 2/88 39 55 77
UDP-1535C - equil Englehard 2/88 43 59 81
US-260 - fresh Englehard 2/88 58 72 90
US-260 - equil Bayway 9/84 45 61 82
US-280E - equil Bayway 3/85 38 59 80
Harshaw/ FSS-1DY - equil Benicia 9/84 38 54 74
Filtrol ROC-1DY - fresh Filtrol 9/86 51 68 90

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 32 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

TABLE 3 (Cont)
FCC CATALYST DENSITY

VENDOR DESIGNATION SOURCE DATE DENSITY-lbs/ft3 (G/CC)

TAPPED PARTICLE SKELETAL

Davison CBZ-1 - fresh Baton Rouge - 35(0.56) - -
CBZ-1 - equil Baton Rouge - 48(0.77) - -
CBZ-1 - equil Baytown - 49(0.79) - -
CBZ-1 - equil Bayway 9/84 53(0.85) 76(1.22) 145(2.32)
DA-250 - fresh - - 52(0.84) - -
DA-250 - equil Davison 6/83 62(0.99) 101(1.62) 157(2.52)
DA-300 - fresh Baton Rouge - 56(0.89) - -
DA-300 - equil Baton Rouge - 59(0.94) - -
DA-400 - fresh - 10/87 55(0.88) 90 (1.44) 149 (2.39)
DA-400 - equil Baton Rouge 9/84 64(1.03) 101(1.62) 169(2.71)
GRZ-1 - fresh Baton Rouge - 45(0.72) - -
GRZ-1 - equil Baton Rouge - 54(0.87) - -
GX-30 - fresh Billings - 44(0.70) - -
GX-30 - equil Billings - 52(0.84) - -
GXO-25 + - fresh Bayway 6/85 50(0.80) 78(1.25) 119(1.91)
GXO-25 + - equil Bayway 10/85 54(0.87) 84(1.35) 142(2.28)
Octacat - fresh - 10/87 45(0.72) 76(1.22) 134(2.15)
RC-25 - fresh Sarnia - 45(0.72) - -
RC-25 - equil Sarnia - 59(0.94) - -
Super D - fresh Billings - 49(0.79) - -
Super D - equil Billings - 57(0.91) - -
Super DX - fresh Billings - 46(0.74) - -
Super DX - equil Billings - 56(0.89) - -
Super DX - equil Benicia 9/84 57(0.91) 88(1.41) 152(2.44)
3B4 - equil Port Jerome 9/74 51(0.82) 84(1.35) 147(2.26)
Engelhard HFZ-20 - equil Engelhard 1/85 50(0.80) 82(1.31) 137(2.20)
HFZ-33 - fresh Engelhard 1/85 54(0.87) 86(1.38) 153(2.45)
HFZ-33 - equil Engelhard 1/85 56(0.90) 91(1.46) 150(2.40)
US-260 - fresh - - 68(1.09) - -
US-260 - equil - - 61(0.97) - -
US-260 - equil Bayway 9/84 63(1.01) 95(1.52) 153(2.45)
US-280 - equil Bayway 3/85 64(1.03) 100(1.60) 160(2.56)
Katalystics BMZ-4 - fresh - - 46(0.74) - -
BMZ-4 - equil - - 49(0.79) - -
Sigma 4 - fresh - - 44(0.70) - -

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 33 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

TABLE 3 (Cont)
FCC CATALYST DENSITY

VENDOR DESIGNATION SOURCE DATE DENSITY-lbs/ft3 (G/CC)

TAPPED PARTICLE SKELETAL

Ketjen KMR-95 - fresh Ingolstadt - 47(0.75) - -
KMR-95 - equil Ingolstadt - 56(0.89) - -
KMR-97 - fresh Ketjen - 46(0.73) - -
KMR-97 - equil Ketjen - 57(0.92) - -
MZ-3 - fresh - - 44(0.70) - -
MZ-7 - fresh Ketjen - 51(0.82) - -
MZ-7 - equil Ketjen - 57(0.91) - -
MZ-7X - fresh Port Jerome - 50(0.80) - -
MZ-7X - equil Port Jerome - 61(0.97) - -
MZ-11 - fresh Ketjen - 49(0.78) - -
MZ-11 - equil Ketjen - 52(0.83) - -
MZ-11X - fresh Ketjen - 47(0.76) - -
MZ-11X - equil Antwerp - 55(0.88) - -
MZ-11P - equil Augusta - 56(0.89) - -
Harshaw/ FSS-1 - fresh - - 47(0.75) - -
Filtrol FSS-1DY - equil Benicia 9/84 52(0.83) 78(1.25) 149(2.39)

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 34 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

FIGURE 1
TERMINAL SETTLING VELOCITY, UTO OR UTCL

100

10

1
1/3
4µ (ρp - ρg)g

3ρg2

0.1
Ut /

0.01

0.001

0.0001
0.1 1 10 100 1000 10,000

ndp

3µ2 1/3
4 ρg (ρp - ρg) g

With n = 1, Ut = Uto
When n ≠ 1, Ut = Utcl
DP6EF01 (n = Cluster Index)

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 35 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

FIGURE 2
EFFECTIVE CLUSTER DIAMETER BASED ON VERTICAL CONVEYING DATA

100
9
8
7
6
5
4
3
Ratio of Cluster Diameter to Single Particle Diameter, n

10
9
8
7
6
5
4
3
ft / s
2

m/s
1.0
9
8
7
6
5
4
3

0.1
0.05 0.1 0.2 0.5 1.0 2 5 10 20 50

DP6EF02 Single Particle Terminal Velocity, ft / s (m / s)

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 36 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

FIGURE 3
SERIES OF TANK MODELS FOR FLUID SOLIDS PROCESS

Bubble
Phase
Gas Inlet

Emulsion
n
Phase

Primary
n
Flow
Solids Inlet

Bypass
m
Flow

Features of Series of Tanks Model

• Each tank assumed to be well mixed, bubble phase assumed to be in or close to plug flow.
• Arrow lines indicate flow of gas or solids between stages. Two way arrows imply additional backmixing.
• Solids are in communication with emulsion and bubble phase tanks (shown as double lines).
• Letters n and m refer to the number of tanks in each line. The larger the number, the closer approximation to plug flow.
• The number of emulsion gas tanks may be assumed to be the same as the number of solids primary flow tanks (as shown).
• Residence time in each line may differ. DP6EF03

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 37 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

FIGURE 4
LOG-NORMAL PROBABILITY PARTICLE SIZE DISTRIBUTION

1000

800

600
500
400

300
Particle Size, dp, (microns)

200

100
dp (50) (56)
80 S= = = 1.5
dp (15.9) (37)
60 dp (50)
50
40
dp (15.9)
30

20

10
0.01 0.1 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.5 99.9 99.99

DP6EF04 Cumulative wt % Less Than Stated Size

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 38 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

FIGURE 5
EFFECTIVE RANGE OF PARTICLE SIZE MEASUREMENT TECHNIQUES

TECHNIQUE TYPICAL PARTICLE SIZE RANGE PARTICLE DIMENSION TYPICAL EQUIPMENT

(microns) DETERMINED

0.001 0.01 0.1 1 10 100 1000 10,000

1. Sieving
Electroformed Minimum Dimension U.S. and Tyler Screens;
Sieves RO-Tap, Fisher-Wheeler,
Standard Screens
Gilson Sieve Shakers;
Electroformed Sieves,
ATM Sonic Sifter, Alpine
Maximum or Minimum Air-Jet Sieve.
Clean Room
Environment and
Electrical Safeguards Standard
2. Electrical
Conductivity Volumetric Coulter Counter, Electro
Equivalent Diameter Zone Celloscope.

3. Microscopy Equivalent Projected Leitz, American Optical,

Area Diameters; Bausch and Lomb,
Electron Microscopy Optical Microscopy Maximum or Minimum Joyce-Loebl, Zeiss Image
Dimension Analyzers or Processors;
Scanning Electron Microscope.

4. Impaction Aerodynamic Andersen, U. Washington,

Equivalent Diameter Brink, Sierra, Scientific Advances
Cascade and Virtual Impactors.
Laboratory Instruments Standard

5. Centrifugation
Stokes' Bahco Particle Classifier,
Equivalent Diameter Joyce Loebl Disc Centrifuge,
Goetz Spiral Centrifuge,
Preining Cylindrical Centrifuge,
MSA (Whitby) Centrifuge.
With Electrostatic Common Practice
6. Elutriation Safeguards and
no Agglomeration Stokes' Roller Particle Size Analyzer
Equivalent Diameter

Gases
7. Sedimentation Stokes' Gas - Sharples Corporation
Equivalent Diameter Micromerograph
Liquids Liquid - Sedigraph

Gases
8. Light Scattering Leeds and Northrup Microtrac,
and Blockage Projected Area
Equivalent Diameter Climet Particle Analyzer,
Liquids Malvern Particle and Droplet
Sizer, HIAC Automatic Particle
Counter, Royco Particle
Counters, Bausch and
Lomb 40 - 1, Cilas Granulometer
DP6EF05
0.001 0.01 0.1 1 10 100 1000 10,000

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-E 39 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

FIGURE 6
ANGLE OF REPOSE FOR GRANULAR SOLIDS

Drained Angle of Repose

β

Funnel

β Poured Angle
of Repose

DP6EF06

FIGURE 7
ANGLE OF INTERNAL FRICTION

Cone of Flowing Solids

Stationary Solids

α = Angle of Internal Friction DP6EF07

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-E 40 of 41
BASIC PROPERTIES OF
FLUID BEDS AND PARTICLES
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

FIGURE 8
DUCTILE AND BRITTLE EROSION

20
Erosion, Wt. per wt. of erodent (x104)

16
Aluminum (Ductile)

12

4 Aluminum Oxide
(Brittle)

0
0 30 60 90

DP6EF08 Angle of Impingement, α

ExxonMobil Research and Engineering Company – Fairfax, VA