ExxonMobil Proprietary

FLUID SOLIDS Section Page

VI-D 1 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

CONTENTS
Section Page

SCOPE ............................................................................................................................................................2

REFERENCES ................................................................................................................................................2

ADDITIONAL LITERATURE ...........................................................................................................................2

GAS DISTRIBUTORS .....................................................................................................................................2

DISTRIBUTOR TYPE..............................................................................................................................2
GRID STABILITY ....................................................................................................................................3
GRID PRESSURE DROP .......................................................................................................................4
GRID JET PENETRATION......................................................................................................................4

HEAT EXCHANGER DESIGN.........................................................................................................................5

INTRODUCTION.....................................................................................................................................5
HEAT TRANSFER BUNDLES.................................................................................................................6
HEAT TRANSFER BETWEEN GAS AND PARTICLES..........................................................................9
TRANSFER LINE COOLERS..................................................................................................................9

SAMPLE CALCULATIONS...........................................................................................................................10
GRID JET PENETRATION....................................................................................................................10
HEAT TRANSFER COEFFICIENTS .....................................................................................................10

COOLING COIL PRELIMINARY DESIGN ....................................................................................................12

NOMENCLATURE ........................................................................................................................................14

TABLES
Table 1 ............................................................................................................................................7
Table 2 ............................................................................................................................................8

FIGURES
Figure 1 Gas Distributor Types........................................................................................................15
Figure 2 Grid Stability ......................................................................................................................16
Figure 3 Effect of Transfer Surface and Fluid Bed Temperature on Bed Emissivity........................17
Figure 4 Heat Transfer Coefficient Maldistribution Factors..............................................................17

Revision Memo
12/02 Corrections and minor updates.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 2 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

SCOPE
This section covers other fluid solids design items such as gas distributors and heat exchangers used to extract heat from
fluidized beds. Distributor types, stability criteria, pressure drop and jet penetration are included. Heat exchanger design and
heat transfer between gas and particles are described.

REFERENCES
1. Shaw, D. F., R. G. Nichols, Refractory Lined Regenerator Grid for Fluid Catalytic Cracking Units, EE.41E.86, April, 1986.
2. Merry, J. M. D., AlChE Journal 21, 507 (1975).
3. Zenz, F. A., Bubble Formation and Grid Design, I Chem E Symposium Series No. 30 (1968).
4. Zilora, S. J., Design Considerations for Steam Generating Coils Located in an FCCU Regenerator, EE.45E.82, May 3, 1982.
5. Design Practices Section IX, Heat Exchange Equipment.
6. Vreedenberg, H. A., Chem. Eng. Science, 9, 52 (1958).
7. Baskakov, A. P., et al., Powder Technology, 8, 273 *73.
8. Exxon Research and Engineering Company, “Industrial Application of Fluidized Bed Combustion, Indirect Fired Process
Heaters, Final Report," U. S. Dept. of Energy, September 30, 1980.
9. Botterill, J. S. M., “Fluid-Bed Heat Transfer," p. 237, Academic Press, London, 1975.
10. Kothari, A. K., M. S. Thesis, Illinois Institute of Technology, Chicago (1967).

ADDITIONAL LITERATURE
1. Baker, C. L. et. al., Fluid Coker Design Notes (Rev. 1), 83 HOP 51, March 14, 1983.
2. Luckenbach, E. C., Design Notes, Catalytic Cracking, Unissued draft report to M. G. Bienstock, Dec. 21, 1989. Available
from the Catalytic Cracking Section, Exxon Engineering.
3. Montiero, F. M., FCCU Maintenance Guide, EE.44E.84, June 1984 with updates.
4. Danielsen, A. M., Flexicoker/Fluid Coker Maintenance Guide, Manual EEEEE-9, December, 1988 with updates.
➧ 5. Mobil FCC Design Guide and Technology Inventory, EE.31E.2002

GAS DISTRIBUTORS

DISTRIBUTOR TYPE
Selection of a gas distributor type is often dependent on specific process characteristics. Several types are shown in Figure 1.
➧ The perforated plate grid (Figure 1a) is probably the most widely used in hExxon commercial scale equipment. It is relatively
cheap and simple to fabricate and maintain, and it should always receive consideration.
For use at very high temperatures, an insulated grid (Figure 1b) may be used. The insulated grid requires that cool gases be
passed up through the grid in order to keep temperatures of structural metal in the plenum at acceptable levels. Grid metal
temperature differentials and stresses are much lower than in unlined grid plates. Three or four in. (75 to 100 mm) of refractory
are placed on the top of the grid. Grid holes are on the order of 3/8 to 1/2 in. (9.5 to 12.5 mm). The grid holes are surrounded by
metal shrouds (pipe stubs) whose length equals the refractory thickness. The diameter of the shrouds is small enough to contain
the expanding gas jet issuing from the grid holes. This is necessary to prevent erosion of the refractory near the gas outlets and
also minimize solids attrition at those points. Erosion and attrition are caused by clockwise recirculating eddies detaching from
the expanding jet, which pull in bed solids and scour the top surface of the grid near the gas outlet holes. The diameter of the
shrouds is calculated from:
H
≥ 5 .2 Eq. (1)
D − do

where: H = The height of the shroud

D = The inside diameter of the shroud
do = The diameter of the grid holes

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 3 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

GAS DISTRIBUTORS (Cont)

H/D ratios of 4:1 are usually used.
Shrouds can also be used to prevent grid erosion and solids attrition in those cases where refractory lining is not required.
Design details are given in Reference (1) and in the FCCU Maintenance Guide.
Dished perforated plates (Figures 1c and 1d) may be used where strength is needed, although present practice is to use flat grids
with a separate support, if possible. Dished plates, especially the upward dished, have an inherent tendency for uneven gas
distribution. This can result in increased pressure drop, grid hole erosion, and solids attrition. This effect can be counteracted by
installing riser pipes so that all gas discharges are at the same elevation (Figure 1e). The grid may be refractory coated on the
bottom to protect against incoming hot gas. The top section may also filled with an insulating refractory. Gas is discharged
through orifices in the caps mounted on the riser pipes.
Bubble cap or tuyere distributors (Figure 1f) are especially useful when it is necessary to prevent backflow of bed particles
through the grid at conditions of low or interrupted flow of fluidizing gas. They are also useful when grid gas jets must enter
downwards into the bed for reasons of fouling.
A pipe grid (Figure 1g) consists of an array of pipes with upward- or downward-facing nozzles. For large diameter vessels, pipe
grids may permit some saving in vessel height when compared to a plate grid with grid plenum beneath. Under certain
conditions mechanical design may be simpler and cheaper, and there is some indication that with downwards-facing nozzles a
lower grid pressure drop is required than for other grid configurations. Pipe grids are especially useful when bed solids are to be
withdrawn downwards through the grid level, as occurs for instance at the bottom of strippers. Gas outlet holes are usually 1/2 to
1 in. diameter (13 to 25 mm) and outlet gas velocity is usually 100 to 150 ft/s (30 to 45 m/s). The holes are shrouded by 2 in.
(50 mm) diameter nipples with a 1:1 to 4:1 length to diameter ratio to minimize solids attrition. Two sets of outlet holes are used,
➧ each angled at approximately 45° from the vertical. A large number of hMobil units use pipe grid distributors and relevant design
information can obtained from EE.31E.2002.

GRID STABILITY
A grid is said to be stable when gas is passing upward through all grid holes and solids are not sifting down through any grid
holes. When this condition occurs, all parts of the bed above the grid will be fluidized, and the bed is also said to be stable. An
unstable bed has slumped or unfluidized zones. Therefore, an important consideration in the design of a large grid is to ensure
sufficient grid pressure drop for good gas distribution and stable bed operation. One approach to calculating grid pressure drop
necessary for stability is shown in Figure 2. The pressure drop through a grid hole will increase with gas velocity. However,
pressure drop through the bed immediately above that hole will decrease as gas preferentially channels through the hole,
because the density of the bed decreases as gas flow increases. Therefore, local grid and bed conditions will move to the
minimum in the total pressure drop curve in Figure 2 by stopping gas flow through some of the holes. In order for the grid to be
stable, total pressure drop through a grid hole plus the bed above the hole must increase as gas flow starts to channel
preferentially. This will halt the channeling by forcing gas to flow through other grid holes as well. A simplified case is instructive.
➧ Assuming that the bed is bubbling freely, pressure drop through the bed is given by:

ρH g ρmf H Eq. (2c)

∆ Pb = = 144 g (psi)
144 gc U−U g
1+ mf c
UB

−3
−3 10 gρ mf H
∆P b = 10 gHρ = (kPa) Eq. (2m)
U − U mf
1+
UB

where: ∆Pb = The pressure drop through the bed (psi), (kPa)
ρ = The fluidized bed density (lb/ft3), (kg/m3)
H = The fluidized bed height (ft), (m)
ρmf = The minimum fluidized bed density (lbs/ft3), (kg/m3)
U = The gas velocity (ft/s), (m/s)
Umf = The minimum fluidization velocity (ft/s), (m/s)
UB = The bubble rise velocity (ft/s), (m/s)
2 2
➧ g = The acceleration due to gravity (32.2 ft/s )(9.81 m/s )
➧ gc = 2
The gravitational conversion factor [(32.2 ft/s ) (lb mass/lb force)]

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 4 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

GAS DISTRIBUTORS (Cont)

and grid pressure drop is:

∆Pg = ρa C2 (Q/A)2 (psi), (kPa) Eq. (3)

where: ∆Pg = The pressure drop through the grid (psi), (kPa)

C = The grid hole coefficient ( s ft ⋅ lb /lbm.), (s m ⋅ kPa

in f kg
Q = The gas flow rate through the grid (ft3/s), (m3/s)
A = The grid hole area (ft2), (m2)
ρa = The density of the mixture flowing through the grid (lb/ft3), (kg/m3)
Solving for the stability condition that d(∆Pg + ∆Pb) / dU > 0 gives the criterion:
∆Pg U
> Eq. (4)
∆Pb 2 (U + UB − Umf )

for stable grid operation. A grid operated at too low a pressure drop will redistribute the gas flow so that some of the holes are
inoperative and the remainder satisfy the criterion. Exxon practice is to maintain grid pressure drops of 25% to 40% of bed
pressure drop to keep all holes active.
Unstable bed conditions are most likely to occur during turndown and certain startup operations when flow rates and grid ∆P are
low. If this occurs, it is usually seen as high temperatures below the grid due to solids sifting, uneven ∆P measurements, and
possibly high solids losses through internal cyclones. The high losses occur if cyclone diplegs terminate in slumped bed zones.
If this happens, the solids in the diplegs back up into the cyclone and make it inoperative. Grid chimneys can be placed on a
small fraction of grid holes to assure that those holes will remain active during low flow conditions. Those holes will remain active
since the bed ∆P above the chimneys is less than that above the remaining grid holes. To prevent cyclone dipleg flooding, it is
recommended that chimneys be placed over half of the grid holes in the vicinity of each dipleg termination. The holes with
chimneys should be alternated with plain holes to minimize re-start problems in the event of loss of fluidizing gas. Chimney ID
should be the same as grid hole ID with a 6:1 chimney length/ID ratio. Chimneys are particularly recommended for grids through
which both gas and solids flow. Startup problems are more likely since only gas is passed through the grid during startup and the
pressure drop is therefore very low. Chimneys can also be used for gas flow only grids, but the need is less than for the gas plus
solids flow design.

➧ GRID PRESSURE DROP

Grid pressure drop is calculated by Equation (3). The grid hole coefficient used in Equation (3) is 0.0135 (conventional units) for
gas flow through a flat grid plate. For grids lined with 4 in. (100 mm) refractory, the coefficient is lower due to pressure recovery
in the 4 in. (100 mm) long grid holes through the refractory section. A coefficient of 0.0115 (conventional units) should be used
for gas flow through refractory lined grids but calculations should be made to test the sensitivity of the grid design, e.g., grid ∆P,
to an assumed coefficient of 0.010 (conventional units). For the case of solids flow through a grid, the coefficient decreases with
hole velocity and solids loading. See Figure 2, Section VI-A for FCC catalyst and gas flow through a grid.
Measurements of grid pressure drop must be corrected for the static head of solids between the upper DP tap and the top of the
grid. Actual flow through the grid must also be used by correcting for any flows which bypass the grid.

GRID JET PENETRATION

Gas entering a fluid bed from an orifice usually seems to form a jet, with jet penetration distances being determined by hole
velocity and diameter. These jets allow gas to bypass the bed and so should not penetrate to the bed surface or occupy an
inordinate fraction of bed height. Jet impingement can also cause severe erosion of surfaces within the bed and such
submerged surface as heat transfer tubes and cyclone diplegs should be located away from any jet penetration zones. An
empirical formula for jet penetration is given by Merry (2):

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 5 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

GAS DISTRIBUTORS (Cont)

æ ρgdo ö 0.3 é æ V 2 ö 0.2 ù

L do = 5.2 ç ÷ ê1.3 ç o ÷ − 1ú Eq. (5)
ç ρpdp ÷ êë ç gdo ÷ úû
è ø è ø
where: L = The jet penetration length (ft), (m)
do = The diameter of a grid hole (ft), (m)
ρg = The density of the fluidizing gas (lbs/ft3), (kg/m3)
ρp = The solids particle density (lbs/ft3), (kg/m3)
dp = The average particle diameter (ft), (m)
Vo = The grid hole gas velocity (ft/s), (m/s)
➧ g = The acceleration due to gravity (32.2 ft/s2), (9.81 m/s2)
Having selected a grid hole velocity and pressure drop sufficient for good gas distribution, significantly higher velocities should be
avoided. Attrition of bed particles increases rapidly as grid hole velocity increases, and erosion of bed internals in the vicinity of
the grid jet can also become serious. However, the effect of grid hole velocity on attrition and erosion can be mitigated by using
grid hole shrouds described previously. High grid pressure drop also wastes compressor power. In some multi-bed processes,
high grid pressure drop adversely affects the pressure balance necessary for solids circulation.
In some cases, grid holes near bed internals may be blanked to prevent jet impingement and erosion of the internals. Since
fluidization in the vicinity of diplegs can affect cyclone performance, the choice of blanking holes to prevent erosion, maintaining
holes, or placing chimneys on holes to promote localized fluidization should be made after considering all issues including those
related to the process.
FCC regenerator grid holes, through which both gas and catalyst pass, have plugged in some units after catalyst changes.
Preliminary findings indicate the plugging is due to the formation of very fine particles resulting from attrition of newer catalysts,
coupled with fluid dynamic effects which tend to cause the fine particles to accumulate near the grid hole wall, and thermal effects
which increase the temperature of theses particles to the softening point. Plugging may be mitigated by using tapered grid holes.
Consult Process Engineering Department for further information.
Penetration from downward flowing jets (e.g., in strippers) is considerably less. An empirical formula for downward jet
penetration is given by Zenz (3):
L / do = 21.6 log Vo ρg − 31.5 Eq. (6c)

L / do = 21.6 log Vo ρg − 33.4 Eq. (6m)

HEAT EXCHANGER DESIGN

INTRODUCTION
In some fluids solids processes, heat must be extracted from a fluidized bed to maintain temperature of the bed within certain
limits or to recover energy from the bed. Examples are in fluidized bed combustion or in FCC regenerators operating under
conditions of high coke yield. Heat exchange was also used in the past in Fluid Hydroforming regenerators. Heat can be
removed by tubular heat exchangers immersed in the fluidized beds or by circulating the solids between the process vessel and
another vessel containing the heat exchanger. The immersed heat exchanger is discussed in this section. It has been used in
Exxon installations in the past, is reliable, but has the disadvantage of not being able to vary the rate of heat removal from the
bed. External heat exchange systems can control the rate of heat removal by controlling the circulation rate. Consult
Engineering Equipment and Environmental Department for further information.
Heat exchanger design technology for fluid solid systems is also available from vendors. UOP has licensed a number of external
dense phase coolers which are reported to be operating satisfactorily. Similar systems are available from Air Products/HRI.
Other vendor systems are also available, but less is known about operating experience and reliability. These include a dilute
phase cooler from Kellogg and a flow through system from UOP. Consult Process Engineering Department if vendor designs are
to be considered.
Immersed heat exchangers are usually cooled by a forced circulation water system. Parallel cooling circuits are used in the
event of tube failure. Other systems which have been used are steam/atomized water and a natural convection circulation
system. The forced circulation system is generally recommended.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 6 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

HEAT EXCHANGER DESIGN (Cont)

HEAT TRANSFER BUNDLES

The heat transfer surface can consist of horizontally or vertically oriented tubes usually containing water and steam as the heat
transfer medium. Other liquids or gases can also be used as the heat transfer medium. However, attention must be given to the
effect of the high heat fluxes associated with heat transfer from a fluidized bed on the heat transfer medium, e.g., coking, vapor
blanketing, etc.
Horizontal tube orientation is generally recommended since design parameters are better defined and vertically oriented tube
bundles were known to be susceptible to vapor locking when used in the past by Exxon in Fluid Hydroformers. Equipment and
Environmental Engineering Department should be consulted if vertical tube orientation is being considered.
Basic design parameters are discussed below. Also see References (4), (5).
Heat Transfer Coefficients. Heat transfer coefficient correlations are available, based on studies published in the literature, for
convective heat transfer to fully immersed horizontal (6) and vertical tube bundles, and for additional radiant heat transfer at
higher temperatures (7). The total heat transfer coefficient, ht, is the sum of the convective, hc, and radiant, hr, coefficients.
Some information is available on heat transfer in the freeboard area, heat transfer to fully immersed extended surface
exchangers (finned tubes) and the variation in local heat transfer coefficients with position on tube surfaces (8). The final point is
important for temperature sensitive heat transfer fluids and is also used to determine maximum tube metal temperatures. These
correlations cover a range of operating variables and can be used for preliminary screening studies. For recommended
correlations to be used in critical designs, consult the Equipment and Environmental Engineering Department.
• Horizontal Tubes. Convective heat transfer coefficients from the fluidized bed to fully immersed horizontal tubes can be
obtained from the Vreedenberg correlations (6).
Gdp ρp
For fine, light particles, i.e., for < 2050 :
ρg µg
−0.3 0.44
é h d ù é 3600 Cpµg ù é Gdto ρp (1 − ε ) ù
ê c to ú ê ú = 0.66 ê ú Eq. (7c)
ëê k g ûú ëê kg ûú ëê ρg µg ε ûú

−0.3 0.44
é h d ù é Cp µg ù é Gdtoρp (1 − ε) ù
ê c to ú ê ú = 0.66 ê ú Eq. (7m)
ëê k g ûú ëê k g ûú ëê ρg µg ε ûú
where: hc = The bed side convective heat transfer coefficient (Btu/hr ft2 °F), (W/m2K)
dto = The tube outside diameter (ft), (m)
Kg = The thermal conductivity of the gas (Btu/hr ft °F), (W/mK)
Cp = The heat capacity of the solids (Btu/lb °F), (J/kgK)
➧ G = The gas superficial mass flux (lbs/ft2s), (kg/m2s)
µg = The gas viscosity (lbs/ft s), (kg/ms)
ε = The bed voidage fraction
ρg = The gas density (lb/ft3), (kg/m3)
ρp = The solids particle density (lb/ft3), (kg/m3)
This equation is claimed to correlate data to within ± 40%.
Gdp ρp
For coarser and heavier particles, i.e., > 2550 :
ρg µg

−0.3 0. 3
éh d ù é Cpµg ù é Gd ρ µ 2g ù
to p
ê c to ú ê3600 ú = 420 ê x 3 2 ú Eq. (8c)
êë k g úû êë k g úû ê ρg µg dp ρp g úû
ë

−0.3 0.3
é h d ù é Cpµg ù é Gd ρ µg2 ù
to p
ê c to ú ê ú = 420 ê x 3 2 ú Eq. (8m)
ëê k g ûú êë k g ûú ê ρg µg dp ρpg ú
ë û

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 7 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

HEAT EXCHANGER DESIGN (Cont)

The correlation [Equation (8)] is claimed to correlate data to within ± 30%. However, these two correlations [Equations (7)
Gdpρp
and (8)] can give substantially different coefficients. For those cases where lies between 2050 and 2550 or where a
ρg µg
series of calculations are being made in which the value moves from less than 2050 to more than 2550, coefficients differing
by as much as a factor of 10 may be calculated when changing from Equation (7) to Equation (8). Caution must be
exercised when using these correlations.
The above correlations were developed for single tubes but can be used for tube bundles. The decrease in heat transfer
coefficients for bundles is only about 10% which is well within the uncertainty range of the correlations.
When the heat transfer tubes are not fully immersed, the heat transfer coefficients are less. Tubes located in the freeboard
region above the bed have heat transfer coefficients lying between those for gas phase bed transfer and for heat transfer to
fully immersed tubes. The coefficients depend on particle concentration at the tube surfaces. Process Engineering
Department should be consulted for specification of heat transfer coefficients for exchangers in the freeboard region above a
fluidized bed.
At temperatures above about 1200°F (920 K), radiation effects become important and the total estimated heat transfer
coefficient must be increased to account for radiation. It is assumed that convective and radiant coefficients are additive.
The effect of radiation is calculated from: (9)
1.71 x 10 −9 εb (Tb4 − Ts4 )
hr = (Btu / hr ft 2 °F) Eq. (9c)
Tb − Ts

5.67 x 10 −8 εb (Tb4 − Ts4 )

hr = ( W / m2 K ) Eq. (9m)
Tb − Ts

where: hr = The radiant heat transfer coefficient (Btu/hr ft2 °F), (W/m2 K)
Tb = The average bed temperature (°R), (K)
Ts = The surface temperature of the tubes (°R), (K)
εb = The bed emissivity
The bed emissivity, εb, is obtained from Figure 3 which relates the emissivity to bed and tube surface temperatures.
Inclusion of radiant heat transfer usually adds about 20% to the value of the convective coefficient for bed temperatures in
the range of 1200 to 1400°F (920 to 1030 K).
For those cases where heat is being removed from the combustion of coke on catalyst in an FCC regenerator and steam is
being generated in the tubes, the following overall heat transfer coefficients, Uo, should be used instead of Equation (7) or
(8).

TABLE 1

CUSTOMARY UNITS

dto - - - - -0.198 - - - - - - - - - -0.375 - - - - - - - - - -0.552 - - - - -

Ug 1 2 3 1 2 3 1 2 3

Uo 69 78 83 58 66 71 49 56 60

METRIC UNITS

dto - - - - -0.060 - - - - - - - - - -0.114 - - - - - - - - - -0.168 - - - - -

Ug 0.305 0.610 0.915 0.305 0.610 0.915 0.305 0.610 0.915

Uo 390 440 470 330 375 400 280 320 340

where: dto = The tube outside diameter (ft), (m)

Ug = The gas superficial velocity (ft/s), (m/s)
Uo = The overall heat transfer coefficient (Btu/hr ft2 °F), (W/m2 K)

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 8 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

HEAT EXCHANGER DESIGN (Cont)

These values which include the effect of radiation, should also be used for preliminary screening purposes only. Consult
Equipment and Environmental Engineering Department for additional information.
Local heat transfer coefficients on the tube surface are not uniform (8). Maximum values of the local coefficients are needed
to calculate the maximum tube metal temperatures and inside film temperatures for temperature sensitive liquids. Local heat
transfer coefficients generally are the highest at the mid-points in the first and fourth quadrants of the tube cross section. In
polar coordinates, this is at 45° and 315° of the cross section. Coefficients at the bottom (180°) are low due to gas
blanketing. Coefficients at the top (0°) are low due to the formation of a cap of dense solids on the top of the tubes. The
points of maximum local heat transfer coefficients correspond to the “surf line" where the dense solids cap ends and the
combination of solids concentration and motion on the tube surface maximizes the rate of heat transfer. Local coefficients
are a function of gas velocity at these points and generally range from 30 to 70% above the average coefficients. This is
shown in Figure 4 where maldistribution factors, h max/h avg, are plotted against the ratio U/Umf. These factors were
measured in beds consisting of large particles with minimum fluidization velocities of about 2 ft/s (0.6 m/s). These effects
have not been verified in beds of smaller particles. Therefore, Figure 4 should be used only for general guidance. Consult
Equipment and Environmental Engineering Department if possible maldistribution of heat transfer coefficients is a critical
issue.
• Vertical Tubes. Heat transfer coefficient correlations for vertical tubes have not been established with the same degree of
reliability as those for horizontal tubes. Generalized correlations available in the literature fit data only to within ± 200%. For
the case where vertical tubes are being used to remove heat from FCC catalyst, the following overall coefficients can be
used for screening purposes.

TABLE 2

CUSTOMARY UNITS METRIC UNITS

Ug 1 2-4 0.305 0.610 - 1.22

Uo 80 88 450 500

Consult Equipment and Environmental Engineering Department for further information.

• Extended Surface Tubes. Heat transfer can be increased by using extended surface (finned) tubes. Although the heat
transfer coefficients based on the total area of the extended surfaces is less than that for bare tubes, the increase in surface
area created by the fins results in an increase in the overall heat transfer rate. The effect of the fins on the heat transfer
coefficients depends on fin size, spacing and orientation (radial vs. axial fins). If extended surface exchangers are being
considered, consult Equipment and Environmental Engineering Department.
Tube Geometry. Horizontal tubes should be arranged in an equilateral triangular pattern with a center to center spacing of 3
tube diameters. This geometry provides adequate solids circulation necessary to maintain high heat transfer coefficients.
Spacing requirements for vertical tubes are not well defined. Consult Equipment and Environmental Engineering Department if
vertical tube orientation is being considered.
Other Design Considerations. The horizontal tubes should be supported by tube sheets with tube sheet spacing based on the
tube length, diameter, metallurgy, and temperature. Minimum tube diameter should be 3 in. (75 mm) to provide sufficient self
support between the tube sheets. Holes should be placed in the sheets to allow horizontal mixing. The holes should be as large
as possible consistent with the structural strength requirements of the tube sheets.
The tube sheets should be supported by legs resting on the grid. The legs should be about 2 ft (0.6 m) high to minimize erosion
due to grid jets and provide good horizontal mixing between the grid and the tube bundles. The grid supports must be designed
to carry the combined load of the tube bundles and the grid.
The implications of a tube failure should be considered and valving added to isolate tubes or groups of tubes in the event of
failure. Instrumentation should be specified in each tube pass to assist in determining if a tube has failed.
Tubes should be designed and operated to prevent phase separation. Short tubing runs and high liquid/vapor weight (L/V) ratios
will minimize phase separation. For the water/steam system, the L/V weight ratio should be 25/1. Velocities in the tubes should
be 20 ft/s (6 m/s) or less.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 9 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

HEAT EXCHANGER DESIGN (Cont)

Design Procedure. The design procedure consists of the following steps:
1. Assume a tube diameter.
2. Calculate the log mean ∆t, based on the average bed temperature and tube interior temperatures.
3. Calculate the bed-side convective heat transfer coefficient hc from Equation (7) or (8). Estimate the radiant heat transfer
coefficient hr from Equation (9). Sum hc and hr to obtain the total bed-side coefficient, hto.
4. Calculate Uo, the overall coefficient from hto, tube wall, fouling factor, and interior heat transfer coefficients or use data from
Table 1 or 2.
5. Calculate the required heat transfer surface area from the required heat transfer rate, Q, Uo and ∆tlm.
6. Calculate the vapor and liquid flow rates at the outlet of the heat exchanger. The vapor rate is determined from the heat
transfer rate and the heat of vaporization. The liquid rate is obtained from the minimum L/V weight ratio (25/1 for
water/steam).
7. Calculate the linear velocities in the heat exchanger and verify that they are below the allowable maximum [20 ft/s (6 m/s) for
water/steam]. If velocities are too high, assume a larger tube diameter and repeat Steps 1 to 7.
8. Lay out the tube bundle(s) based on the required length, diameter [3 in. (75 mm) minimum] and an equilateral triangular
arrangement with a center-to-center spacing of 3 tubing diameters. Tube bundle location should not cause interference with
solids movement, circulation, sprays, diplegs, etc. Determine the number of bundles based on piping, support, isolation
valving, maintenance, header design and other considerations.
9. Calculate the effects of plant start-up, shutdown, upsets and turndown on the performance of the heat exchangers. Verify
that the exchangers will operate satisfactorily under these conditions.

HEAT TRANSFER BETWEEN GAS AND PARTICLES

Heat transfer rates between gas and particles in a fluidized bed are high due to the intimate contact between gas and particles.
This results in nearly isothermal bed conditions for most operations. A correlation is available (10) which gives the heat transfer
coefficient between gas and particles:
1.3
hp dp é dp ρg Ug ù
= 0.03 ê ú Eq. (10)
kg êë µg úû

where: hp = The gas to particle heat transfer coefficient (Btu/hr ft2 °F), (W/m2 K)
dp = The weight mean particle diameter (ft), (m)
kg = The thermal conductivity of the gas (Btu/hr ft °F), (W/mK)
ρg = The gas density (lb/ft3), (kg/m3)
Ug = Gas superficial velocity (ft/s), (m/s)
µg = Gas viscosity (lb/ft s), (kg/ms)
The correlation is good to within about ± 50%

TRANSFER LINE COOLERS

Cooling in pneumatic transfer lines can be accomplished using natural convection cooling aided by external fins. The current
basis for estimating the inside heat transfer coefficient is to multiply the coefficient for flowing air by 2. Standard correlations can
be used to obtain the air only inside coefficient. The external coefficient can be obtained from standard correlations for the
natural convection cooling of finned tubes. The inside coefficient estimated in this fashion is conservative. Caution should be
used if overcooling can result in condensation.

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 10 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

SAMPLE CALCULATIONS

GRID JET PENETRATION

• Calculate upward jet penetration length, L, using Equation (5)
Given: do = 0.042 ft
ρg = 0.44 lbs/ft3
ρp = 101 lbs/ft3
dp = 1.84 x 10-4ft
Vo = 100 ft/s

é ρg do ù
0.3 é éV2 ù
0.2 ù
L ê1.3
= 5. 2 ê ú ê o ú − 1ú
ê ú
do ëê ρp dp ûú êë
êë g do úû
úû

é 0.44 x 0.042 ù
0.3
é é 100 2 ù
0.2
ù
= 5.2 ê −4 ú
ê1.3 ê ú − 1ú
ë101 x 1.84 x 10 û êë ë 32.2 x 0.042 û úû

= 35

L = 35 x 0.042 = 1.5 ft

• Calculate downward jet penetration length, L, using Equation (6)

L
= 21.6 log Vo ρg − 31.5
do

= 21.6 log 100 0.44 − 31.5

= 7.85
L = 7.85 x 0.042 = 0.33 ft

HEAT TRANSFER COEFFICIENTS

Given: dto = 0.375 ft
kg = 0.04 Btu/hr ft °F
Cp = 0.2 Btu/lb °F
µg = 2.5 x 10-5 lbs/ft s
G = 0.24 lb/ft2 s
ρp = 101 lbs/ft3
ρg = 0.07 lbs/ft3
dp = 1.84 x 10-4 ft

• Calculate
G dp ρp 0.24 x 1.84 x 10 −4 x 101
=
ρg µg 0.07 x 2.5 x 10 −5

= 2600

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 11 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

SAMPLE CALCULATIONS (Cont)

• Calculate bed side convective heat transfer coefficient hc, using Equation (8), since above value > 2550
− 0. 3 0.3
éh d ù é c p µg ù éG d ρ µg2 ù
to p
ê c to ú ê3600 ú = 420 ê x 3 2 ú
êë k g úû êë k g úû ê ρg µg dρ ρρ g ú
ë û

− 0.3
é hc x 0.375 ù é 0.2 x 2.5 x 10 −5 ù
ê ú ê3600 ú
ë 0.04 û êë 0.04 úû

( )
0.3
é 0.24 x 0.375 x 101 2.5 x 10 −5
2
ù
= 420 ê x ú
êë 0.04 x 2.5 x 10
−5
1.84 x 10 −
(4 3 2
)
(101) x 32.2 úû
hc = 324 Btu/hr ft2 °F
Decrease hc by 10% to account for bundling effect
hc = 324 x 0.9 = 292 Btu/hr ft2 °F

• Calculate radiant heat transfer coefficient, hr, using Equation (9)

Given: Tb = 1760°R
Ts = 1260°R
εb = 0.6 (Figure 3)

hr =
(
1.71 x 10 −9 εb Tb 4 − Ts 4 )
Tb − Ts

=
(
1.71 x 10 −9 x 0.6 1760 4 − 1260 4 )
1760 − 1260

= 15 Btu/hr ft2 °F

• Calculate total outside coefficient, hto

hto = hc + hr

= 292 + 15 = 307 Btu/hr ft2 °F

• Calculate gas to particle heat transfer coefficient, hp, using Equation (10)
Given: U = 3.5 ft/s
1.3
hp dp é dp ρ g U ù
= 0.03 ê ú
kg êë µg úû

1. 3
hp x 1.84 x 10 −4 é 1.84 x 10 − 4 x 0.07 x 3.5 ù
= 0.03 ê ú
0.04 êë 2.5 x 10 −5 úû

hp = 14 Btu /hr ft2 °F

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 12 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

COOLING COIL PRELIMINARY DESIGN

• Prepare preliminary design of heat transfer coil for removal of heat from fluid bed. Use horizontal coils with steam
generation in coils. Estimate coil surface area, internal velocity.
Given: Q = 4 x 106 Btu/hr
tbed = 1300°F
Cooling water inlet temperature = 100°F
Saturated steam produced at 600 psig (490°F)
Ug = 3 ft/s
1. Assume 4 in. nominal coil pipe size
dto = 0.375 ft
2. Calculate

∆tlm =
(1300 − 100 ) − (1300 − 490 )
ln
(1300 − 100 )
(1300 − 490 )
= 990°F
3. Obtain Uo
From Table 1, Uo = 71 Btu/hr ft2
If hto were calculated from Equations (8) and (9), using properties of particles and solids from previous samples, and using
1000, 600, and 200 for inside coefficient, wall coefficient and fouling factor, respectively, Uo would be about 90 Btu/hr ft2 °F,
about 25% greater.
Use Uo = 71 Btu/hr ft2 °F for conservative estimate.
4. Calculate heat transfer surface area, coil length
Q
A =
Uo ∆tlm

4 x 10 −6
=
71 x 990

= 57 ft2

57 ft 2
L =
1.047 ft 2 / ft

= 54 ft
5. Estimate internal velocity
Btu 1
Vapor rate = 4 x 106 x
hr 871 Btu / lb steam

= 4590 lbs/hr
Liquid rate for 25/1 L/V weight ratio
= 25 x 4590
= 115,000 lbs/hr

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 13 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

COOLING COIL PRELIMINARY DESIGN (Cont)

Vapor specific volume = 0.770 ft3/lb (steam tables)
Liquid specific volume = 0.0201 ft3/lb (steam tables)
Volumetric flow rate
Vapor = 4590 x 0.770 = 3530 ft3 / hr
Liquid = 115,000 x 0.0210 = 2310 ft3 / hr
Total = 5840 ft3 / hr

ft 3 hr
Velocity = 5840 x x 0.0884 ft 2
hr 3600 s

= 18 ft/s
Since velocity is less than 20 ft/s, design is satisfactory.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 14 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

NOMENCLATURE
A = Grid hole area, ft2 (m2)
æ ö
➧ C = ç lb
çs ft f /in.÷÷ (s kPa m/kg)
ç lbm ÷
è ø
Cp = Heat capacity of solids, Btu/lb°F (J/kg K)
D = Shroud inside diameter, ft (m)
do = Grid hole diameter, ft (m)
dp = Average particle diameter, ft (m)
dto = Heat transfer tube outlet diameter, ft (m)
G = Gas or solids mass flux, lbs/ft2s (kg/m2s)
➧ g = Acceleration due to gravity, (32.2 ft/s2)(9.81 m/s2)
H = Fluidized bed height, shroud height, ft (m)
hc = Bed side convective heat transfer coefficient, Btu/hr ft2°F (W/m2K)
hp = Gas to particle heat transfer coefficient, Btu/hr ft2°F (W/m2K)
hr = Radiant heat transfer coefficient, Btu/hr ft2°F (W/m2K)
hto = Sum of bed side and radiant heat transfer coefficients, Btu/hr ft2°F (W/m2K)
kg = Thermal conductivity of gas, Btu/hr ft°F (W/mK)
L = Grid jet penetration length, ft (m)
L/V = Liquid to vapor ratio
Q = Gas flow rate through grid, ft3/s (m3s)
Tb = Average bed temperature, °R (K)
Ts = Surface temperature of heat transfer tubes, °R (K)
U, Ug = Superficial gas velocity, ft/s (m/s)
UB = Bubble rise velocity, ft/s (m/s)
Umf = Superficial gas velocity at minimum fluidization, ft/s (m/s)
Uo = Overall heat transfer coefficient, Btu/hr ft2°F (W/m2K)
Vo = Grid hole velocity, ft/s (m/s)
∆Pb = Bed pressure difference, lbs/in.2 (kPa)
∆Pg = Grid pressure difference, lbs/in.2 (kPa)
∆t = Temperature difference °F (°C)
∆tlm = Log mean temperature difference, °F (°C)
ε = Voidage fraction
εb = Bed emissivity
µg = Gas viscosity, lb/ft s (kg/ms)
ρ = Fluid bed density, lbs/ft3 (kg/m3)
ρa = Density of mixture flowing through grid, lbs/ft3 (kg/m3)
ρg = Gas density, lb/ft3 (kg/m3)
ρmf = Fluid bed density at minimum fluidization, lbs/ft3 (kg/m3)
ρp = Solids particle density, lbs/ft3 (kg/m3)

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 15 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

FIGURE 1
GAS DISTRIBUTOR TYPES

a. Perforated Plate b. Refractory Lined Grid

D
Refractory

c. Upward Dish d. Downward Dish

e. Refractory Lined Dish

f. Bubble Cap Distributor
with Riser Pipes

g. Pipe Grid

DP6DF01

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page FLUID SOLIDS
VI-D 16 of 17
OTHER ITEMS
December, 2002 PROPRIETARY INFORMATION - For Authorized Company Use Only DESIGN PRACTICES

FIGURE 2
GRID STABILITY

Umf

Total Pressure Drop

Unstable Stable
Pressure Drop

Bed Pressure Drop

Grid Pressure Drop

DP6DF02 Fluidizing Gas Velocity

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
FLUID SOLIDS Section Page
VI-D 17 of 17
OTHER ITEMS
DESIGN PRACTICES PROPRIETARY INFORMATION - For Authorized Company Use Only December, 2002

FIGURE 3
EFFECT OF TRANSFER SURFACE AND FLUID BED TEMPERATURE ON BED EMISSIVITY

0.9
600°C
1110°F
700°C
1290°F
0.8 800°C
1470°F
900°C
1650°F
Bed Emissivity, Eb

0.7
1000°C
1830°F

0.6

0.5
Lines of Constant
Fluid Bed Temperature
0.4

0.3
300 400 500 600 700 800 900 1000 (°C)
570 750 930 1110 1290 1470 1650 1830 (°F)

DP6DF03 Transfer Surface Temperature

FIGURE 4
HEAT TRANSFER COEFFICIENT MALDISTRIBUTION FACTORS

2.0
Maldistribution Factor

1.5
Bed Temperature Range 525 – 1550°F
(275 – 840°C)

1.0

Maldistribution Factor = hmax/ have

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

DP6DF04 Ratio of Superficial Gas Velocity to Superficial Minimum Fluidization Velocity, U/Umf

ExxonMobil Research and Engineering Company – Fairfax, VA