262 DRYERS AND COOLING TOWERS

TABLE 9.12. Performance of Drum Dryers

(a) Single, Double drum and Vacuum Drums

~~

Moisture Content,
(% Wet Basis) Steam Drum Feed Capacity
Method Pressure. Speed Temp. [Ib product/ Vacuum
Material of Feed Feed Product (Ib/sq in.) (rpm) (“F) (hr)(sqft)] (in. Hg)

Double-drum dryer
Sodium sulfonate trough 53.6 6.4 63 8; 164 7.75
Sodium sulfate trough 76.0 0.06 56 7 150 3.08
Sodium phosphate trough 57.0 0.9 90 9 180 8.23
Sodium acetate trough 39.5 0.44 70 3 205 1.51
Sodium acetate trough 40.5 10.03 67 8 200 5.16
Sodium acetate trough 63.5 9.53 67 8 170 3.26
Single-drum dryer
Chromium sulfate spray film 48.5 5.47 50 5 - 3.69
Chromium sulfate dip 48.0 8.06 50 4 - 1.30
Chromium sulfate pan 59.5 5.26 24 2; 158 1.53
Chromium sulfate splash 59.5 4.93 55 1; 150 2.31
Chromium sulfate splash 59.5 5.35 53 4% 154 3.76
Chromium sulfate dip 59.5 4.57 53 5; 153 3.36
Vegetable glue pan 60-70 10-12 20-30 6-7 - 1-1.6
Calcium arsenate slurry 75-77 0.5-1.0 45-50 3-4 - 2-3
Calcium carbonate slurry 70 0.5 45 2-3 - 1.5-3
Twin-drum dryer
Sodium sulfate dip 76 0.85 55 7 110 3.54
Sodium sulfate top 69 0.14 60 9; 162 4.27
Sodium sulfate top 69 5.47 32 9; 116 3.56
Sodium sulfate splash 71 0.10 60 6 130 4.30
Sodium sulfate splash 71.5 0.17 60 12 140 5.35
Sodium sulfate splash 71.5 0.09 60 10 145 5.33
Sodium phosphate splash 52.5 0.59 58 5; 208 8.69
Sodium phosphate dip 55 0.77 60 5; 200 6.05
Sodium sulfonate top 53.5 8-10 63 a; 172 10.43
Vacuum single-drum dryer
Extract pan 59 7.75 35 8 - 4.76 27.9
Extract pan 59 2.76 35 6 - 1.92 27.9
Extract pan 59 2.09 36 4 - 1.01 atmos.
Extract pan 56.5 1.95 35 7; - 3.19 22.7
Extract Pan 56.5 1.16 50 2; - 0.75 atmos.
Skim milk Pan 65 2-3 10-12 4-5 - 2.5-3.2
Malted milk pan 60 2 30-35 4- 5 - 2.6
Coffee pan 65 2-3 5-10 1-1; - 1.6-2.1
Malt extract spray film 65 3-4 3-5 0.5-1.0 - 1.3-1.6
Tanning extract pan 50-55 8-10 30-35 8-10 - 5.3-6.4
Vegetable glue pan 60-70 10-12 15-30 5-7 - 2-4
(Perrys Chemical Engineers Handbook, McGraw-Hill, 1950 edition).

calculation of residence time when assumptions are made about
terminal temperatures.
9.9. FLUIDIZED BED DRYERS
Free flowing granular materials that require relatively short drying
times are particularly suited to fluidized bed drying. When longer
drying times are necessary, multistaging, recirculation or batch
operation of fluidized beds still may have advantages over other
modes.
A fluidized bed is made up of a mass of particles buoyed up out
of permanent contact with each other by a flowing fluid. Turbulent
activity in such a bed promotes high rates of heat and mass transfer
and uniformity of temperature and composition throughout. The
basic system includes a solids feeding device, the fluidizing chamber
with a perforated distributing plate for the gas, an overflow duct for
removal of the dry product, a cyclone and other equipment for
collecting fines, and a heater and blower for the gaseous drying
medium.
Much ingenuity has been applied to the design of fluidized bed
drying. Many different arrangements of equipment are illustrated
and described in the comprehensive book of Kro11 (1978) for
instance. Figure 9.13(a) depicts the basic kind of unit and the other
items are a few of the many variants. Tables 9.14 and 9.15 are
selected performance data.
Shallow beds are easier to maintain in stable fluidization and of
course exert a smaller load on the air blower. Pressure drop in the
air distributor is approximately l p s i and that through the bed
equals the weight of the bed per unit cross section. Some pressure
drop data are shown in Table 9.14. The cross section is determined
by the gas velocity needed for fluidization as will be described. It is
usual to allow 3-6 ft of clear height between the top of the bed and
the air exhaust duct. Fines that are entrained are collected in a
cyclone and blended with the main stream since they are very dry
 9.9. FLUIDIZED BED DRYERS 263
TABLE 9.12-(continued)

(b) Single and Double Drum with Various Feed Arrangements

Moisture
Content Vapor
Pressure Rotation Unit Product Drying
out Absolute S p e Rate
Kind of Dryer,
Kind of Stock
In
(%) (%I (bar) (l/min) (fiJ3r) (kg/m2 hr)

Single drum, dip feed

Alkali carbonate 50 8 bis 12 3.5 4.4 20 17.8
Double drum, dip feed
Organic salt solution 73 2.8 5.5 5 6.8 18.6
Organic compound, dilute slurry 70 1.2 5.5 5 8.6 19.6
Organic compound, solution 75 0.5 5.0 5 1.1 1.9
Single drum with spreading rolls
Skim milk concentrate 50 4 3.8 24 15.8 14.2
Whey concentrate 45 4.3 5.0 16 10 bis 11.8 7.4 bis 8.8
Cuprous oxide 58 0.5 5.2 10 11.0 14.3
Single drum, splash feed
Magnesium hydroxide, dense slurry 65 0.5 3.0 1 6.8 5.4
Double drum, splash feed
Iron hydroxide, dilute slurry 78 3.0 3.0 3 15.4 4.7
Organic salt, dilute slurry 80 1.7 bis 3.1 5.0 3 bis 5 3.6 bis 6.8 13.3 bis 26.2
Sodium acetate 50 4.0 6.0 5 10.0 9.3
Sodium sulfate 70 2.3 7.8 5 18.0 40.4
Double drum, top feed
Beer yeast 80 8.0 6.0 5 10.0 36.2
Skim milk, fresh 91.2 4.0 6.4 12 6.2 61.5
Organic salt solution 89 - 6.0 5.5 4 32.3
Organic salt solution 60 3 5 bis 6 6.5 12.2 17.7
Organic compound, dilute slurry 75 1 3.5 4.5 1.4 bis 6.8 12.6 bis 18
Double drum with spreading rolls
Potato DUID 76.2 11.4 8 5 22.5 61.1

(Kroll, 1978, p. 348).

due to their small size. Normally entrainment is 5-10% but can be
higher if the size distribution is very wide. It is not regarded as
feasible to permit high entrainment and recycle back to the drying
chamber, although this is common practice in the operation of
catalytic cracking equipment.
Mixing in shallow beds is essentially complete; Figure 9.5(c)
shows some test data in confirmation. The corresponding wide
distribution of residence times can result in nonuniform drying, an
effect that is accentuated by the presence of a wide distribution of
particle sizes. Multiple beds in series assure more nearly constant
residence time for all particles and consequently more nearly
uniform drying. The data of Table 9.14(b) are for multiple zone
dryers. Figures 9.13(c) and (d) have additional zones for cooling the
product before it leaves the equipment. Another way of assuring

TABLE 9.13. PerformanceData of Pneumatic Conveying Dryers (Sketches in Fig. 9.12)

(a) Raymond Flash Dryer

Fine Spent Organic Chicken Fine Coal

Mineral Grain Chemical Droppings Filter Cake

Method of feed pump belt screw Pump screw

Material size, mesh -100 - -30 - -30
Product rate (Ib/hr) 27,000 9000 900 2300 2000
Initial moisture content (% w/w basis) 25 60 37 70 30
Final moisture content (% w/w basis) nil 12 3 12 8.5
Air inlet temperature F‘F) 1200 1200 450 1300 1200
Air outlet temperature PF) 200/300 200/300 200/300 200/300 200/300
Method of heating direct direct direct direct direct
oil oil oil oil oil
Heat consumption (Btu/lb water evaporated) 1.6 x l o 3 1.9 x l o 3 3.1 x io3 1.9 x i o 3 1 . 4 i~o 3
Air recirculation no no no no no
Material recirculation yes yes no yes no
Material of construction MS MS/SS MS MS MS
Fan capacity (std. ft3/min) 18,000 22,000 4300 8500 1500
Installed fan HP 110 180 30 50 10
Product exit temperature (“F) 200 - 200 - 135
(Courtesy International Combustion Products Ltd.; Williams-Gardner, 1971).
 264 DRYERS AND COOLING TOWERS
TABLE 9.13-( continued)

(b) Buttner-Rosin Pneumatic Dryer

Metallic Adipic Coal Filter

Stearate Starch Acid Fiber Cake

Method of feed sling sling screw distributor distributor

Material size fine fine -30 mesh -a1 .in. -30 mesh
Product rate (Ib/hr) 280 13,236 10,000 2610 67,200
Initial moisture (% w/w basis) 40 34 10 62.4 32
Final moisture (% w/w basis) 0.5 13 0.2 10 6
Air inlet temperature ("F) 284 302 320 752 1292
Air outlet temperature ( O F ) 130 122 149 230 212
Method of heating steam steam steam oil PF
Heat consumption (Btu/lb/water evaporated) 2170 1825 2400 1720 1590
Air recirculation no no no no yes
Material recirculation yes no yes yes yes
Fan capacity (std. ft3/min) 1440 26,500 9500 12,500 27,000
Installed fan HP 15 220 65 60 250
Product exit temperature ("F) 104 95 120 140 158
(Courtesy Rosin Engineering Ltd.; Williams-Gardner, 1971).

(c) Pennsalt-Berks Ring Dryer

Metal" Spent" Sewageb Polystyrene

Stearates Grains Sludge Starches Beads

belt back vibratory cascading vibratory

feeder mixer feeder rotary feeder
rotary rotary rotary valve rotary
Method of Feed valve valve valve screen valve

Product rate (Ib/hr) 240 1120 4300 5000 1000

Initial moisture (% w / w basis) 55 80 45 35 2.0
Final moisture (% w/w basis) 1 5 12 10 0.2
Air inlet temperature ("F) 250 500 600 300 175
Air outlet temperature ( O F ) 150 170 170 130 115
Method of heating steam gas oil steam steam
Heat consumption (Btu/lb water evaporated) 2900 1800 1750 2000 5000
Air recirculation no no no no no
Material recirculation yes yes yes no no
Material of construction ss MSG MS MSG ss
Fan capacity (std ft3/min) 3750 16,500 8250 15,000 900
Installed fan HP 20 75 60 60 7.5

Ring dryer application.

(Courtesy Pennsalt Ltd.; Williams-Gardner, 1971).

(d) Various Pneumatic Dryers

~~~ ~

Air/Solid
Gas Ratio
Tube Tube Rate Gas Temp ('c) Solid Solid Temp ('c)Moisture (%) (m3/kg) Evaporated
Water
Dia Height (m3/ hr) Rate
Material Location (cm) (m) (NTP) In Out (kg/hr) In Out In Out (NTP) (kg/kg) (kg/hr)

Ammonium
sulphate Japan 18 1 1100 215 76 950 38.5 63 2.75 0.28 1.2 1.5 23.5
Sewage sludge
filter cake U.S.A. - 1200 700 121 2270 15 71 80 10 5.3 7.2 1590
Coal 6 m m U.S.A. - 50,000 371 80 51,000 15 57 9 3 1.o 1.3 4350
Hexamethylene
tetramine Germanv 30 38" 3600 93 50 2500 - 48 6-10 0.08-0.15 1.4 1.9 18.1

"23 m vertical, 15 m horizontal.

(Nonhebel and Moss, 1971).
 9.9. FLUIDIZED BED DRYERS 265

Recirculation

Expansion
bellows

Change over flop

t
b l e paddle mixer

H a m m e r mill

B o t t o m bend
-.-.-. .---
Combustion chamber

(a) (b)

1. Fan
2. Ring duct
3. Manifold
4. Inlector
5 . Air outlet
6. F e e d e r
7. Filter
8. H e a t e r
9. Cyclone 7
IO. Disintegrator
II. Bag filter
12. Discharge

Figure 9.12. Examples of pneumatic conveying dryers; corresponding performance data are in Table 9.13. (a) Raymond flash dryer, with a
hammer mill for disintegrating the feed and with partial recycle of product (Raymond Division, Combustion Engineering). (b) Buttner-Rosin
pneumatic dryer with separate recycle and disintegration of large particles (Rosin Engineering Ltd.). (c) Berks ring dryer; the material
circulates through the ring-shaped path, product is withdrawn through the cyclone and bag filter (Penmalt Chemical Co.).

complete drying is a recirculation scheme like that of Figure
9.13(e). In batch operation the time can be made as long as
necessary.
Stable fluideation requires a distribution of particle sizes,
preferably in the range of a few hundred microns. Normally a size
of 4mm or so is considered an upper limit, but the coal dryers of
Tables 9.15(a) and (b) accommodate sizes up to 0.5in. Large and
uniformly sized particles, such as grains, are dried successfully in
spouted beds [Fig. 9.13(f)]: Here a high velocity gas stream entrains
the solid upward at the axis and releases it at the top for flow back
through the annulus. Some operations do without the mechanical
draft tube shown but employ a naturally formed central channel.
One way of drying solutions or pastes under fluidizing
conditions is that of Figure 9.13(g). Here the fluidized mass is of
auxiliary spheres, commonly of plastic such as polypropylene, into
which the solution is sprayed. The feed material deposits uniformly
on the spheres, dries there, and then is knocked off automatically as
it leaves the drier and leaves the auxiliary spheres behind. When a
mass of dry particles can be provided to start a fluidized bed drying
process, solutions or pastes can be dried after deposition on the
seed material as on the auxiliary spheres. Such a process is
employed, for instance, for growing fertilizer granules of desired
larger sizes, and has largely replaced rotary dryers for this purpose.
A few performance data of batch fluid dryers are in Table
266 DRYERS AND COOLING TOWERS

EXAMPLE 9.8 At a tower diameter of 0.6 m,

Suing a Pneumatic Conveying Dryer
A granular solid has a moisture content of 0.035 kg/kg dry material
which is to be reduced to 0.001 kg/kg. The charge is at the rate of
u=-= Q [25.0 m/sec at 450"C,
0.36n/4 13.6 m/sec at 120°C.
9.72 kg/sec, is at 60°C and may not be heated above 90°C. Inlet air
is at 450°C and has a moisture content of 0.013 kg/kg dry air. These velocities are great enough to carry the largest particles with
settling velocity of 10 m/sec.
Equations are developed over intervals in which Wl-+ W,,

v-
Tl+ T,, and Ti+ Ti.
The procedure will be:
1. Start with known W,, T,, and T;.
g2 2. Specify a moisture content W,.
T2= 85 F 3. Assume a value T2 of the solid temperature.
4. Calculate Ti from the heat balance.
90 F 5. Check the correctness of T, by noting if the times for heat and
-T2, w2 mass transfers in the interval are equal.

m, = 9.72 kgls Heat balance:

7
T, = 60 C
fis[0.391(T,- T,)+(W,-WZ)(T,- T,+563)]
9 W, = 0.035 kg/kg
=fii,,{[O.25 +0.48(0.001](T; - Ti)
Specific gravity of the solid is 1.77 and its heat capacity is + 0.48(W,- W,)(T; - 60)).
0.39 cal/g "C. The settling velocity of the largest particle present,
2.5 mm dia, is 10m/sec. Heat capacity of the air is taken as Substitute fis/fia= 9.72/3.46 = 2.81 and solve for Ti.
0.25 cal/g "C and the latent heat at 60°C as 563 cal/g. Experimental
data for this system are reported by Nonhebel and Moss (1971, pp. T; =
240Q and are represented by the expressions: -0.25048T; + 28.8(W1 - W,) + 2.81
Heat transfer coefficient: X[O.39(T'-T1)+(W1-Wz)(T,-T1+563)]
ha = 0.47 cal/(kg solid)("C).
0.48( W, - W,) - 0.25048
Vapor pressure:
(1)
-
P = exp(13.7419- 5237.0/T), atm, K.
g,=0.013+~(Wl-0.013)=0.013+2.81(W, -0.013). (2)
ma
Mass transfer coefficient:
p - g1 g1
(partial pressure in air).
kga = exp(-3.1811- 1.7388In w - 0.2553(1n w)', - 18/29 + g, = 0.6207 g,+ (3)

where w is the moisture content of the solid (kg/kg) in the units kg

+
g, = 0.013 2.81(W2- 0.013). (4)
water/(kg solid)(atm)(sec). p - g2

In view of the strong dependence of the mass transfer - 0.6207 + g,

coefficient on moisture content and the 35-fold range of that Pal = exp[13.7419- 5237.9/(T1+ 273.2)], vapor pressure. (6)
property, the required residence time and other conditions will be Pa, = exp[13.7419 - 5237.9/(T2+ 273.2)]. (7)
found by analyzing the performance over small decrements of the
moisture content.
An air rate is selected on the assumption that the exit of the
T;-T,-(Ti-T,)
solid is at 85°C and that of the air is 120°C. These temperatures
(AT)""=In[(T;- &)/(Ti- T,)]'
need not be realized exactly, as long as the moisture content of the
exit air is below saturation and corresponds to a partial pressure less AQ=O.391(T,-T,)+(W,-Wz)(T,- T1+563),
than the vapor pressure of the liquid on the solid. The amount of per kg of solid. (10)
heat transferred equals the sum of the sensible heat of the wet solid = 0.5( W, + W,). (11)
and the latent heat of the lost moisture. The enthalpy balance is k,u = exp[-3.1811- 1.7388 In W - 0.2553(1n W)']. (12)
based on water evaporating at 60°C:
e,, = AQ/hu(AT),, = AQ/O.43(AT),,, heating time. (13)
+
fis[(0.39 0.001)(85 - 60) + (0.035 - 0.001)(85 - 60 + 563)] 0, = (W, - W,)/kga(AP),,, mass transfer time. (14)
+
= fi,[(0.25 + 0.480(0.001))(450 - 120) 0.48(0.034)(120 - 60)], 2 = 0, - 0, +0 when the correct value of T,
has been selected. (15)
After the correct value of Tz has been found for a particular
interval, make W,+ W,, T2+Tl, and T$-+Ti. Specify a
 9.9. FLUIDIZED BED DRYERS 267

EXAMPLE
9.&(continued) 38 I N P U T W l , W Z , T l , A l ! A 1 is th
c inlet air temp T 1 '
decremented value of W,, assume a value of T,, and proceed. The
48 I N P U T T 2 ! T r i a l value
solution is tabulated. 5 8 A2=(2.51*(.391t<T2-Tl}+CWl-W
W T T' e w Z>X(T2-T1+563?>-.2584~*A1+23
.8*~Wl-W2})/(.48bCWl-wz)-.25
0.035 60 450 0 048 >
0.0325 73.04 378.2 0.0402 68 G1=.813+2.82t(Wl-.813>
0.03 75.66 352.2 0.0581
0.025 77.41 31 5.3 0.0872
78 Pl=G1/(.6207+Gl)
0.02 77.23 286.7 0.1133
80 G2=.013+2.31S(W2-.813}
0.015 76.28 261.3 0.1396 90 PZ=G2/(.6287+G2>
0.01 75.15 236.4 0.1687 108 Ql=EXPCl3.7419-5237.9~~Tl+27
0.005 74.67 208.4 0.2067 3 . 2 > j ! vapor pressure
0.003 75.55 192.4 0.2317 1 1 8 Q2=EXPC13.7419-5237.9,(T2+27
0.001 79.00 165.0 0.2841 3.2))
1 2 8 P3=CQl-Pl-Q2+P2?/LOG<CQl-Pl>
When going directly from 0.035 to 0.001, /CQ2-P2)) ! C a P > l m
130 T 3 = ( A l - T l - H 2 + T 2 ? ~ L O C ( ( A 1 - T 1 >
T2= 80.28, /(RZ-T2i> ! taT}lm
T ; = 144.04, 1 4 8 Q=.391tCTZ-T1>+CWl-W2jX<T2-T
1 +563 j
8 = 0.3279 sec. 158 H l = Q / . 4 7 / T 3 ! heatins time
The calculation could be repeated with a smaller air rate in order to 168 W = . 5 S C W 1 + W 2 ~
reduce its exit temperature to nearer 120"C, thus improving thermal 1 7 8 K=EXP(-3.1811-1.7388XLCIG<W>-
efficiency. .2533XLOG<W}*2)
In the vessel with diameter = 0.6 m, the air velocities are 188 H 2 = ( W l - W 2 ) / K / P 3 ! vaporizati
25.0 m/sec at 450°C inlet
on time
198 Z=Hl-H2 ! time difference s h
5.15 m/sec at 165°C outlet ould be zero
288 D I S P Z
20.1 m/sec average. 2 1 4 DISP A 2 , H l
228 GOTO 4 0 ! i f Z is n o t rlear e
The vessel height that will provide the needed residence time is noush to zero.; otherwise the
correct value o f T 2 ha5 bee
H = &e = 20.1(0.2841) = 5.70 m. n found
Very fine particles with zero slip velocity will have the same 2 3 0 END
holdup time as the air. The coarsest with settling velocity of
10 m/sec will have a net forward velocity of
iis = 20.1 - 10 = 10.1m/sec,
which corresponds to a holdup time of
e = s . 7 / i o . i =0.56sec,
Data f o r t h e f i r s t i n t e r v a l
which is desirable since they dry more slowly.
After the assumption of T2, other quantities are evaluated in
the order shown in this program.

1 8 ! Example 9 . 3 . P n e u m a ? ic cor1
verins drrer
28 ! Findins the exit solids ? e
m P T 2 b r trial, then a l l de^
endent quanti t ies

9.14(a). This process is faster and much less labor-intensive than
tray drying and has largely replaced tray drying in the pharma-
ceutical industry which deals with small production rates. Drying
rates of 2-l0lb/(hr)(cuft) are reported in this table, with drying
times of a fraction of an hour to several hours. In the continuous
operations of Table 9.15, the residence times are at most a few minutes.
Thermal efficiency of fluidized bed dryers is superior to that of
many other types, generally less than twice the latent heat of the
water evaporated being required as heat input. Power requirements
are a major cost factor. The easily dried materials of Table 9.15(a)
show evaporation rates of 58-103 lb/(hr)(HP installed) but the
more difficult materials of Table 9.15(d) show only 5-18 Ib/(hr) (HP
installed). The relatively large power requirements of fluidized bed
dryers are counterbalanced by their greater mechanical simplicity
and lower floor space requirements.
Air rates in Table 9.15 range from 13 to 793 SCFM/sqft, which
is hardly a guide to the selection of an air rate for a particular case.
A gas velocity twice the minimum fluidization velocity may be taken
as a safe prescription. None of the published correlations of
minimum fluidizing velocity is of high accuracy. The equation of
Leva (Fluidization,McGraw-Hill, New York, 1959) appears to be
as good as any of the later ones. It is
(9.20)
268 DRYERS AND COOLING TOWERS
where G,,,, is in lb/(hr)(sqft), pg and ps are densities of the gas and
solid (Ib/cuft), D, is the particle diameter (in.), and p is the gas
viscosity (cP). In view of the wide scatter of the data on which this
correlation is based, shown on Figure 6.14(f), it appears advisable
to find the fluidization velocity experimentally for the case in hand.
Although it is embarrassing again to admit the fact,
unfortunately all aspects of fluidized bed drying must be established
with pilot plant tests. The wide ranges of performance parameters
in Tables 9.14 and 9.15 certainly emphasize this conclusion. A
limited exploration of air rates and equipment size can be made on
the basis of a drying rate equation and fluidization correlations from
the literature. This is done in Example 9.9. A rough approximation
of a drying rate equation can be based on through circulation drying
of the granular material on a tray, with gas flow downward.
9.10. SPRAY DRYERS
Suitable feeds to a spray dryer are solutions or pumpable pastes and
slurries. Such a material is atomized in a nozzle or spray wheel,
contacted with heated air or flue gas and conveyed out of the
equipment with a pneumatic or mechanical type of conveyor.
Collection of fines with a cyclone separator or filter is a major
aspect of spray dryer operation. Typical equipment arrangements
and flow patterns are shown in Figure 9.14.
The action of a high speed spray wheel is represented by Figure
9.14(e); the throw is lateral so that a large diameter vessel is required
with this form of atomization, as shown in Figure 9.14(a). The flow
from nozzles is largely downward so that the dryer is slimmer and
taller. Parallel flow of air and spray downward is the most common

cyclone

inle 1 + DV material

Figure 9.13. Fluidized bed dryers. (a) Basic equipment arrangement (McCabe and Smith, Unit Operations in Chemical
Engineering, McGraw-Hill, New York, 2984). (b) Multiple bed dryer with dualflow distributors; performance data are
in Table 9.14(b) (Romankou, in Dauidson and Harrison, Fluidisation, Academic, New York, 2972). (c) A two-bed
dryer with the lower one used as cooler: (a, b, c) rotary valves; (d) drying bed; (e) cooling bed; (f, g) air distributors;
(h, i) air blowers; (k) air filter; (I) air heater; (m) overflow pipe; (n) product collector (Kroll, 2978). (d) Horizontal
multizone dryer: (a) feeder; (b) air distributor; (c) fluidized bed; (d) partitions; (e) dust guard; (f) solids exit; (8) drying
zone; (h) cooling zone; (i, k) blowers; (1, m) air plenums; (n) air duct; (0)dust collector; (p) exhaust fan (Kroll, 2978).
(e) Circulating fluidized bed used for removal of combined water from aluminum hydroxide: (a) feed; (b) fluidized bed;
(c) solids exit; (d) fuel oil inlet; (e) primary air inlet; ( f ) secondary air inlet; (8) gas exit (Kroll, 2978). ( f ) Spouted bed
with draft tube for drying coarse, uniform-sized granular materials such as grains [Yang and Keairns, AIChE Symp.
Ser. 176, 218 (2978),Fig. I].(g) Fluidized bed dryer for sludges and pastes. The fluidized solids are fine spheres of
materials such as polypropylene. The wet material is sprayed in, deposits on the spheres and dries there. At the outlet
the spheres strike a plate where the dried material is knocked off and leaves the dryer as flakes. The auxiliary spheres
remain in the equipment: (a) feed; (b) distributor; (c) spheres loaded with wet material; (d) returning spheres; (e)
striking plate; (f) hot air inlet; (g) air and solids exit (Kroll, 2978).
 9.10. SPRAY DRYERS 269

Gas Out
A

Alternative
Solids feed

i -Solid flow
Gas and Solids Feed --- Gas flow
(e) (R
Figure 9.13-(continued)

arrangement, but the left-hand figure of Figure 9.14(d) is in
counterflow. Figure 9.14(c) has tangential input of cooling air. In
some operations, the heated air is introduced tangentially; then the
process is called mixed flow. Most of the entries in Table 9.16(a) are
parallel flow; but the heavy duty detergent is in counterflow, and
titanium dioxide is either parallel or mixed flow. Counterflow is
thermally more efficient, results in less expansion of the product
particles, but may be harmful to thermally sensitive products
because they are exposed to high air temperatures as they leave the
dryer. The flat bottomed dryer of Figure 9.14(c) contacts the exiting
solids with cooling air and is thus adapted to thermally sensitive
materials.
Two main characteristics of spray drying are the short drying
time and the porosity and small, rounded particles of product. Short
270 DRYERS AND COOLING TOWERS
TABLE 9.14. Performance Data of Fluidized Bed Dryers: Batch and Multistage Equipment

(a) Batch Dryers

Lactose
Ammonium Base Pharmaceutical Liver Weed
Bromide Granules Crystals Residue Killer
Holding capacity (Ib wet product) 100 104 160 280 250
Bulk density, dry (lb/ft3) 75 30 20 30 35
Initial moisture (% w/w basis) 6 10 65 50 20-25
Final moisture (% w/w basis) 1 2 0.4 5.0 1.o
Final drying temperature ("F) 212 158 248 140 140
Drying time (min) 20 90 120 75 210
Fan capacity (ft3/min at 11 in. w.g.) 750 1500 3000 4000 3000
Fan HP 5 10 20 25 20
EvaDoration rate (Ib H,O/hr) 15 5.7 52 100 17
(Courtesy Calmic Engineering Co. Ltd.; Williams-Gardner, 1971).

(b) Multistage Dryers with Dual-flow Distributors [Equipment Sketch in Fig. 9.13(b)]

Function Heater Cooler Drier Cooler

Wheat Wheat Quartz
Material Grains Grains Slag Sand
Particle size (diameter)(mm) 5x3 5x3 0.95 1.4
Material feed rate (metrictons/hr) 1.5 1.5 7.0 4.0
Column diameter (m) 0.90 0.83 1.60 1.70
Perforated trays (shelves):
Hole diameter (mm) 20 20 20; 10 20
Proportion of active section 0.4 0.4 0.4; 0.4 0.4
Number of trays 10 6 1: 2 20
Distance between trays (mm) 20 20 25; 40 15
Total pressure drop on fluidized bed (kgf/m2) 113 64 70' 40
Hydraulic resistance of material on one tray (kgf/rnz) 7.8 9.2 20; 10 1.8
Inlet gas temperature ("C) 265 38 300 20
Gas inlet velocity (m/sec) 8.02 3.22 4.60 0.74
Material inlet temperature PC) 68 175 20 350
Material discharge temperature ("C) 175 54 170 22
Initial humidity (% on wet material) 25 - 8 -
Final humidity ( % on wet material) 2.8 - 0.5 -
Blower conditions
Pressure (kgf/mz) 450 250 420 250
Throughput (m3/min) 180 130 360 100
(80°C) (50°C) (70°C) (35°C)
Power consumption (HP) 50 20 75 7.5

'With grids and two distributor plates.

(Romankov,in Davidson and Harrison, Nuidisation, Academic, New York, 1971).

drying time is a particular advantage with heat sensitive materials. Residence times of air and particles are far from uniform; Figure
Porosity and small size are desirable when the material sub- 9.5(a) and (b) is a sample of such data.
sequently is to be dissolved (as foods or detergents) or dispersed Because of slip and turbulence, the average residence times of
(as pigments, inks, etc.). Table 9.17 has some data on size particles are substantially greater than the mean time of the air,
distributions, bulk density, and power requirements of the several definitely so in the case of countercurrent or mixed flow. Surface
types of atomizers. moisture is removed rapidly, in less than 5 sec as a rule, but falling
The mean residence time of the gas in a spray dryer is the ratio rate drying takes much longer. Nevertheless, the usual drying
of vessel volume to the volumetric flow rate. These statements are operation is completed in 5-30 sec. The residence time distribution
made in the literature regarding residence times for spray drying: of particles is dependent on the mixing behavior and on the size
distribution. The coarsest particles fall most rapidly and take
Source Time (sec) longest for complete drying. If the material is heat-sensitive, very
tall towers in parallel flow must be employed; otherwise,
Heat Exchanger Design Handbook (1983) 5-60 countercurrent or mixed flows with high air temperatures may
McCormick (1979) 20 suffice. In some cases it may be feasible to follow up incomplete
Masters (1976) 20-40 (parallel flow)
spray drying with a pneumatic dryer.
Nonhebel and Moss (1971) <60
Peck (1983) 5-30
Drying must be essentially completed in the straight sided
Wentz and Thygeson (1979) zones of Figures 9.14(a) and (b). The conical section is for gather-
Williams-Gardner (1971) 4-10 (<15ftdia) ing and efficient discharge of the dried product. The lateral throw
10-20 b15ftdia) of spray wheels requires a vessel of large diameter to avoid
 9.10. SPRAY DRYERS 271

TABLE 9.15. Performance Data of Continuous Fluidized Bed Dryers

(a) Data of Fluosatatic Ltd.

Coal Sand Silica Sand Limestone Iron Ore

Material size, mesh ;-0 -250 - 18-0 &-0 -2-0

Method of feed twin bucket conv. conv. conv.
screw elev.
Product rate (Ib product/hr) 448,000 22,400 112,000 67,000 896,000
Initial moisture (% w / w basis) 11 6 6 15 3
Final moisture (% w / w basis) 5.5 0.1 0.1 0.1 0.75
Residence time (min) 1 1.25 1.5 1.25 0.5
Dryer diameter (ft) 10 3.0 7.25 5.5 8.5
Fluid bed height (in.) 18 12 12 12 18
Air inlet temperature (“F) 1000 1200 1200 1200 1200
Air outlet temperature (“F) 170 212 212 212 212
Air quantity (ft3/min std.) 40,000 2000 9000 13,000 45,000
Material exit temperature (“F) 140 220 220 220 220
Evaporation (Ib/hr) 24,640 1430 6720 11,880 20,400
Method of heating coal gas oil oil oil
Heat consumption (Btu/lb water evaporated) 1830 1620 1730 1220 2300
Fan installed HP 240 20 80 115 350

(Williams-Gardner, 1971).

(b) Data of Head Wrightston Stockton Ltd.

Silicious Glass
Coal Grit Sand Sand Asphalt

Method of feed screw chute chute chute chute

feeder
1. 1 . 1 . 3 .
Material size - 3 in. -=in. -36 mesh -E in. -iij in.
Product rate (Ib product/hr) 190,000 17,920 15,680 33,600 22,400
Initial moisture (% w/w basis) 14 5 7 5 5
Final moisture (% w/w basis) 7 0 0 0 0.5
Residence time (min) 2 1; 3 3 10
Dryer diameter 7ft3in. 3ftOin. 4ft6in. 6ft6in. 8ftOin.
Fluid bed height (in.) 21 12 12 12 24
Air inlet temperature (“F) 1000 1400 1400 1400 470
Air outlet temperature ( O F ) 135 230 230 230 220
Air quantity (ft3/min std) 20,000 2000 2000 3500 7000
Material exit temp (“F) 140 230 230 230 220
Evaporated rate (Ib/hr) 11,200 896 1097 1680 1120
Method of heating coke- gas oil town gas oil gas oil
oven gas
gas
Heat consumption (Btu/lb water evaporated) 2000 2250 2000 2200 1800
Fan installed HP 210 321 18 30 90

(Williams-Gardner, 1971).

(c) Data of Pennsalt Ltd.

Abrasive Clay Granular Household

Grit Granules Sand Desiccant Salt

Product rate (Ib/hr) 2200 1000 14,000 150 13,500

Initial moisture (% w/w basis) 9 22 6 25 4
Final moisture (% w/w basis) dv 3 dry 7 0.03
Air inlet temperature (“F) 580 160 325 300 390
Air outlet temperature (OF) 210 120 140 205 230
Method of heating gas steam gas gas steam
Heat consumption (Btu/lb water evaporated) 2700 3800 2700 3600 5100
Bulk density (lb/ft3) 120 60 90 30 60
Average drying time (min) 2.5 30 3 24 4
Fan capacity (ft3/min std.) 2.5 1.35 1.05 0.84 1.05
Installed fan HP 10 45 25 5 50
(Williams-Gardner, 1971).
272 DRYERS AND COOLING TOWERS
TABLE g.l&(continued)
(d) Data of Rosin Engineering Ltd.

Sodium Weed
Perborate Killer PVC Coal Sand

Method of feed screw vibrator screw vibrator vibrator

Material size 30-200 5-1 m m 60- 120 3 mesh- 30-1 20
mesh flake mesh zero mesh
Product rate (Ib product/hr) 11,400 5100 10,075 440,000 112,000
Initial moisture (% w/w basis) 3.5 14 2.0 8 8
Final moisture (% w/w basis) 0.0 0.2 0.2 1 0.2
Residence time (min) 1.5 11 30 0.3 0.45
Drier bed size (hx ft) 22.5 X 5.5 18 X 4.5 23 x 6 16 x 6.6 12.5 x 3.2
Fluid bed height (in.) 4 3 18 5 6
Air inlet temperature (OF) 176 212 167 932 1202
Air outlet temperature ("F) 104 150 122 180 221
Air quantity (ft3/min std) 6600 14,200 5400 67,330 8000
Material exit temperature ("F) 104 205 122 180 212
Evaporation (Ib/hr) 400 720 183 33,440 9750
Method of heating steam steam steam coke-oven oil
gas
Heat consumption (Btu/lb water evaporated) 21 00 3060 4640 1970 2200
Fan installed HP 33 40 34 600 70

(Williams-Gardner, 1971).

EXAMPLE 9.9 - 688(0. O0787)"83[0.048( 150 - 0.048)]0.94

S i g a Fluidized Bed Dryer (0.023)0.88
A wet solid at 100°F contains W = 0.3 Ib water/lb dry and is to be
= 17.17 Ib/(hr)(sqft).
dried to W = 0.01. Its feed rate is 100 Ib/hr dry. The air is at 350°F
and has Hg0=0.015Ib water/lb dry. The rate of drying is
represented by the equation Let Gf = 2Gmf = 34.34 Ib/(hr)(sqft).
Expanded bed ratio
dW
-_= 60(Hs - H,), (Ib/lb)/min. (L/Lo)= (Gf/Gm,)o.22
= 2O." = 1.16.
d6
Take voidage at minimum fluidization as
The solid has a heat capacity 0.35 Btu/(lb)("F), density 150 lb/cuft,
and average particle size 0.2pm (0.00787in.). The air has a
, 0.40,
E ~ =
viscosity of 0.023 CP and a density of 0.048 Ib/cuft. The fluidized
bed may be taken as a uniform mixture. A suitable air rate and :. Ef = 0.464.
dimensions of the bed will be found:
Drying time:
I -
Wo - W 0.3 - 0.01
e = 60(Hs - H,) = 60(Hs - H,) '

Since complete mixing is assumed, H, and H, are exit

conditions of the fluidized bed.
Humidity balance:
T, = 100 F (T,)
-,I A(H, - Hgo)= S( Wo - W ) ,
H, = 0.015 + 0.29S/A.

Average heat capacity:

C, = f(C,, + C,) = 0.24 + 0.45[(0.015 + H,)/2]

Symbols used in the computer program are in parentheses. = 0.2434 + O.225Hg. (3)

Minimum fluidizing rate by Leva's formula: Heat balance:

6880~83[0.048(150- 0.048)]0.94 AC,(Tg, - Tg) = S[(C, + W ) ( T - To)+ d(W0 - W)l,

Gmf =
Po.88 (A/S)C,(350 - Tg) = 0.36(T - 100) + W(0.29). (4)
 9.10. SPRAY DRYERS 273
EXAMPLE
9.9--(continued) When drying is entirely in the falling rate period with rate equation
Adiabatic saturation line:

(5)
the drying time will be
Vapor pressure:

P, = exp[11.9176- 7173.9/(T, + 389.5)]. (6)

Saturation humidity: where Hs, Hgr and W are final conditions. When the final W is
small, 0.01 in the present numerical example, the single stage drying
time will be prohibitive. In such cases, multistaging, batch drying,
(7) or some other kind of drying equipment must be resorted to.

Eliminate T3 between Eqs. (4) and (5):

1 8 ! Example 9 . 3 . F l u i d i z e d Ped
0.36( T4- 100) + 261 d rY e r
T,=350-
RCg 2 8 I N P U T R ! =HfS, r a t i o cat r a t
e 5 l3t f l o w of a i r a n d s o l i d
39 H 3 = . 8 1 5 + 2 9 / R ! =Hs
49 C1=.2434+.225tH3
59 INPUT T 4 ! T r i a l v a l u e o f T s
Procedure: For a specified value of R = A I S , solve Eqs. (6), 68 LOSUB 299
(7),and (8) simultaneously. 78 '{ 1=y
s9 T 4 = 1 . 0 8 9 1 t T 4
99 GOSUB 299
R T, r, H, 4 tl (min) 1 9 8 '{ 2 = y
5 145.14 119.84 0.0730 0.0803 0.662 118 K = . 0 0 0 l t Y l / ( Y Z - Y l ~
6 178.11 119.74 0.0633 0.0800 0.289 128 T4=T4/1.8891-K
8 220.09 119.60 0.0513 0.0797 0.170 139 DISP T4
10 245.72 119.52 0.0440 0.0795 0.136 1 4 9 I F H B S < K . / T 4 j < = . 6 9 8 0 1 THEN 1 6
12 262.98 119.47 0.0392 0.0794 0.120 0
1 5 8 GOTO 68
Take 1 6 0 D I S P USING 1 7 8 j R , T Z , T 4 > H 3 >
H4 T 5#.

1 7 8 IMRGE D D , X I D D U . D , X , D D D . D , X , .
R = 10 Ib air/lb solid, DUD0 > x , .DDDD,>SI . D D D
A = lO(100) = lo00 Ib/hr, 188 E HD
0 = 0.136 min. 298 ! SR f o r T 4
2 1 8 P = .E..X P ( 1 1 . 9 1 7 6 - 7 1 7 3 . 9 . ( T 4 + J S 9
. J > *'
Cross section: 228 H 4 = 1 8 S P y 2 9 / < 1 - P > ! = H.3
258 T 3 = T 4 + 9 8 8 % ( H 4 - H 3 > / C l ! = T 4
A/G, = 1000/34.34 = 29.12 sqft, 6.09 ft dia. E48 Y = - T 3 + 3 5 9 - ( . 3 6 t ( T 4 - 1 0 8 ) + 2 6 1 )
/ R. ./ e - 1
-
Avg density: 2'59 T 5 = . 2 9 i { H 4 - H 3 > / 6 9 ! = time
268 RETURN
$(1/20.96 + 1/19.03) = 0.0501 Ib/cuft. 279 EtJD
Linear velocity:

u=-- Gf - 34.34
= 24.62 fpm
ps(60) 0.0501(0.464)(60) T. H I1 Time
Bed depth:
-+
R
Ts A S -
5 1 4 5 . 1 1 1 3 . 8 4 ,97338 ,61883 662
L = U B = 24.62(0.136) = 3.35 ft.
6 1 7 8 . 1 1 1 9 . 7 4 .8633 .8888 ,289
8 2213.1 1 1 9 . 6 1 , 4 5 1 3 ,8797 ,178
la 2 4 5 . 7 1 1 9 . 5 3 , 9 4 4 8 .879rJ ,136
Note: In a completely mixed fluidized bed, the drying time is 1 2 2 6 3 . 8 1 1 9 . 4 7 .a392 ,8794 ,128
determined by the final moisture contents of the air and solid. 15 2 8 8 . 4 1 1 9 . 4 2 , 8 3 4 3 .8792 ,188
274 DRYERS AND COOLING TOWERS
A i r diffuser
Exhaust o ~ r
plates

AIR 230
FjED
I 6o I
,Feed

rj

+c--

productto
collectw
Sweeper air

Figure 9.14. Spray dryer arrangements and behavior. (a) Spray dryer equipped with spray wheel; straight section LID = 0.5-1.0 (Proctor and
Schwartz Znc.). (b) Spray dryer equipped with spray nozzle; straight section LID = 4-5 (Nonhebel and Moss, 1971). (c) Spray dryer for very
heat sensitive products; flat bottom, side air ports and air sweeper to cool leaving particles. (d) Distribution of air temperatures in parallel
and countercurrent flows (Musters, 1976,p . 18, Fig. 1.5).(e) Droplet-forming action of a spray wheel (Stork-Bowen Engineering Co.).
 9.10. SPRAY DRYERS 275

TABLE 9.16. PerformanceData of Spray Dryers

(a) Data of Kroll(1978)

~

Moisture Air
Content Temperature
spry Flow
Kind of Stock In(%) Out(%) Device Pattern In (“C) Out (“C)

Skim milk, d = 6 0 p m 48-55 4 wheel or parallel 250 95-100

50-60 4 nozzle
170-200 bar parallel 250 95-100
Whole milk 50-60 2.5 wheel or
nozzle
100-140 bar parallel 170-200
Eggs, whole 74-76 2-4 wheel or
nozzle parallel 140-200 50-80
Eggs, yolks 50-55 2-4 wheel or
nozzle parallel 140-200 50-80
Eggs, whites 87-90 7-9 wheel or
nozzle parallel 140-200 50-80
Coffee, instant, 300 p m 75-85 3-3.5 nozzle parallel 270 110
Tea, instant 60 2 nozzle,
27 bar parallel 190-250
Tomatoes 65-75 3-3.5 wheel parallel 140-1 50
Food yeast 76-78 8 wheel parallel 300-350 100
Tannin 50-55 4 wheel parallel 250 90
PVC emulsion, 90% > 80 pm 40-70 0.01-0.1 wheel or
<60pm nozzle or
pneumatic parallel 165-300
Melamine-urethane-formaldehyde resins 30-50 0 wheel
140-160 m/sec parallel 200-275 65-75
Heavy duty detergents 35-50 8-13 nozzle,
30-60 bar counter 350-400 90-1 10
Kaolin 35-40 1 wheel parallel 600 120

(b) Performance of a Dryer 18ft Dia by 18ft High with a Spray Wheel and a Fan Capacity of 11,000cfm at the
Outleta

Air Temp (“F)

96 Water Evaporation
Material In out in Feed Rate (Ib/hr)
~ ~~~

Blood, animal 330 160 65 780

Yeast 440 140 86 1080
Zinc sulfate 620 230 55 1320
Lignin 400 195 63 910
Aluminum hydroxide 600 130 93 2560
Silica gel 600 170 95 2225
Magnesium carbonate 600 120 92 2400
Tanning extract 330 150 46 680
Coffee extract A 300 180 70 500
Coffee extract B 500 240 47 735
Magnesium chloride 810 305 53 1140
(to dihydrate)
Detergent A 450 250 50 660
Detergent B 460 240 63 820
Detergent C 450 250 40 340
Manganese sulfate 600 290 50 720
Aluminum sulfate 290 170 70 230
Urea resin A 500 180 60 505
Urea resin 8 450 190 70 250
Sodium sulfide 440 150 50 270
Pigment 470 140 73 1750
‘The fan on this dryer handles about 11,00Ocuft/min at outlet conditions. The outlet-air temperature includes cold air
in-leakage, and the true temperature drop caused by evaporation must therefore be estimated from a heat balance.
(Bowen Engineering Inc.).
276 DRYERS AND COOLING TOWERS
TABLE 9.17. Particle Diameters, Densities, and Energy orifice diameters are 0.012-0.15 in. An acceptably narrow range of
Requirements droplet sizes can be made for a feed of particular physical properties
by adjustment of pressure and diameter. Multiple nozzles are used
(a) Atomizer Performance
for atomization in large diameter towers. Because of the expense of
Power Input motive air or steam, pneumatic nozzles are used mostly in small
TVPe Size Range (pm) (kWh/1000 L) installations such as pilot plants, but they are most suitable for
dispersion of stringy materials such as polymers and fibers. The
Single fluid nozzle 8-800 0.3-0.5 droplet size increases as the motive pressure is lessened, the range
Pneumatic nozzle 3-250
of 60-1OOpsi being usual. The action of a rotating wheel is
Spray wheel 2-550 0.8-1 .O
Rotating cup 25-950 indicated in Figure 9.14(e). Many different shapes of orifices and
vanes are used for feeds of various viscosities, erosiveness, and
clogging tendencies. Operating conditions are up to 60,OOO Ib/hr per
(b) Dry Product Size Range atomizer, speeds up to 20,000rpm, and peripheral speeds of
250-600 ft/sec.
Product pm The main variables in the operation of atomizers are feed
Skim milk 20-250 pressure, orifice diameter, flow rate and motive pressure for nozzles
Coffee 50-600 and geometry and rotation speed of wheels. Enough is known about
Eggs 5-500 these factors to enable prediction of size distribution and throw
Egg white 1-40 of droplets in specific equipment. Effects of some atomizer
Color pigments 1-50 characteristics and other operating variables on spray dryer
Detergents 20-2000 performance are summarized in Table 9.18. A detailed survey of
Ceramics 15-500 theory, design and performance of atomizers is made by Masters
(c) Bulk Density of Sprayed Product as Affected by Air Inlet (1976), but the conclusion is that experience and pilot plant work
Temperature and Solids Content of Feed" still are essential guides to selection of atomizers. A clear choice
between nozzles and spray wheels is rarely possible and may be
arbitrary. Milk dryers in the United States, for example, are
W t % s o l i d s in feed
equipped with nozzles, but those in Europe usually with spray
wheels. Pneumatic nozzles may be favored for polymeric solutions,
although data for PVC emulsions in Table 9.16(a) show that spray
wheels and pressure nozzles also are used. Both pressure nozzles
and spray wheels are shown to be in use for several of the
applications of Table 9.16(a).

APPLICATIONS
For direct drying of liquids, slurries, and pastes, drum dryers are
the only competition for spray dryers, although fluidized bed dryers
sometimes can be adapted to the purpose. Spray dryers are capable
of large evaporation rates, 12,000-15,000Ib/hr or so, whereas a
300sqft drum dryer for instance may have a capacity of only
3000 Ib/hr. The spherelike sprayed particles often are preferable to
drum dryer flakes. Dust control is intrinsic to spray dryer

0
100
d
200
I
A
300
-
400
I I
500
I
600
construction but will be an extra for drum dryers. The completely
enclosed operation of spray dryers also is an advantage when toxic
or noxious materials are handled.
A i r inlet, OC
THERMAL EFFICIENCY
'The full lines are against temperature, the dashed ones against Exit air usually is maintained far from saturated with moisture and
concentration: (a) sodium silicate; (b) coffee extract, 22%; (c) water at a high temperature in order to prevent recondensation of moisture
dispersible dye, 19.5%; (d) gelatin. in parallel current operation, with a consequent lowering of thermal
[Data of Duffie and Marshall, Chem. Eng. Prog. 49, 417 480 (195311.
efficiency. With steam heating of air the overall efficiency is about
40%. Direct fired dryers may have efficiencies of 8 0 4 5 % with inlet
temperatures of 500-550°C and outlet of 65-70°C. Steam
accumulation of wet material on the walls; length to diameter ratios consumption of spray dryers may be 1.2-1.8 Ib steam/lb evapor-
of 0.5-1.0 are used in such cases. The downward throw of nozzles ated, but the small unit of Table 9.19(b) is naturally less efficient. A
permits small diameters but greater depths for a given residence 10% heat loss through the walls of the dryer often is taken for
time; LID ratios of 4-5 or more are used. design purposes. Pressure drop in a dryer is 15-50in. of water,
depending on duct sizes and the kind of separation equipment used.

ATOMIZATION
Proper atomization of feed is the key to successful spray drying.
The three devices of commercial value are pressure nozzles,
pneumatic nozzles, and rotating wheels of various designs. Usual
pressures employed in nozzles range from 300 to 4000psi, and
DESIGN
The design of spray dryers is based on experience and pilot plant
determinations of residence time, air conditions, and air flow rate.
Example 9.10 utilizes such data for the sizing of a commercial scale
spray dryer.
 9.11. THEORY OF AIR-WATER INTERACTION IN PACKED TOWERS 277

TABLE 9.18. Effects of Variables on Operation of Spray Dryers

Variable Increased Factors Increased Factors Decreased

Chamber inlet temperature Feed rate and thus: bulk density ( b )

Chamber outlet temperature
Gas volume rate
Feed concentration
Atomizer speed
Atomizer disc diameter
For stable lattices
For unstable lattices
Atomizer vane depth
Atomizer vane number
Atomizer vane radial length
Feed surface tension
Chamber inlet gas humidity
product rate,
particle size ( b ) ,
product moisture content,
chamber wall build-up (a)
product thermal degradation feed rate and thus:
(a) product rate
particle size ( b )
product moisture content
chamber wall build-up
feed rate and thus: residence time
product rate,
particle size ( b ) ,
product moisture content,
chamber wall build-up (a)
product rate,
bulk density (b),
particle size ( b )
bulk density particle size and thus:
product moisture content
chamber wall build-up
coagulation (a) and thus:
particle size,
product moisture content,
chamber wall build-up
bulk density ( b ) particle size ( b )and thus:
product moisture content,
chamber wall build-up
For unstable lattices
particle size
chamber wall build-up
bulk density ( b ) particle size ( b )
product moisture content,
chamber wall build-up (a)

aThis factor will only occur if a critical value of the variable is exceeded.
Not for suspensions.
(Nonhebel and Moss, 1971).

The smallest pilot unit supplied by Bowen Engineering has a Analysis of the interaction of air and water involves the making
diameter of 30 in. and straight side of 29 in., employs parallel flow, of material and enthalpy balances. These are made over a
up to 25ACFM, 150-1000”F, particle sizes 30-40pm average, differential section of the tower shown on Figure 9.15(a) and are
either pneumatic nozzle or spray wheel. The performance of this subsequently integrated to establish the size of equipment for a
unit is given in Table 9.19. The magnitude of the “product number” given performance. In terms of empirical heat, k,, and mass, k,,
is arrived at by pilot plant work and experience; it increases with transfer coefficients, these balances are
increased difficulty of drying or thermal sensitivity or both.
Although much useful information can be obtained on this small Gdh = LC, dT = L dT (9.21)
scale, Williams-Gardner (1971) states that data on at least a 7 ft dia = k,(hs- h ) dz (9.22)
dryer be obtained for final design of large capacity units. = kh(T - T,) dz. (9.23)

9.1 1. THEORY OF AIR-WATER INTERACTION IN PACKED In Eq. (9.21) the heat capacity of water has been taken as unity.
TOWERS The approximations that are involved in making an enthalpy
difference a driving force are discussed for example by Foust et al.
The key properties of mixtures of air and water vapor are described (1980). Rearrangement and integration leads to the results
in Section 9.1. Here the interactions of air and water in packed
towers under steady flow conditions will be analyzed. The primary
objectives of such operations may be to humidify or dehumidify the (9.24)
L
air as needed for particular drying processes or other processes, or
to cool process water used for heat transfer elsewhere in the plant. dT
Humidification-dehumidification usually is accomplished in spray (9.25)
towers, whereas cooling towers almost invariably are filled with
some type of packing of open structure to improve contacting but _- (9.26)
with minimum pressure drop of air.
278 DRYERS AND COOLING TOWERS
TABLE 9.19. Product Numbers and Performance of a 30 x Both forms of the integral are employed in the literature to define
29 in. Pilot Plant Spray Dryer the number of transfer units. The relation between them is
(a) Product Numbers of Selected Materials
k,Z/G = (L/G)(NTU). (9.27)
Material Product nunher The height of a transfer unit is

HTU = Z/(NTU)= L / k , = ( L / G ) ( G / k m ) . (9.28)

I. COLOURS
Reactive dyes 5- 6 The quantity G / k , sometimes is called the height of a transfer unit
Pigments 5-11 expressed in terms of enthalpy driving force, as in Figure 9.16, for
Dispersed dyes 16-26 example:
2. FOODSTUFFS
Carbohydrates 14-20
G / k , = (G/L)(HTU). (9.29)
Milk 17
Integration of Eq. (9.21) provides the enthalpy balance around
Proteins 16-28
one end of the tower,
3. PHARllACEUTICALS
L(T - Tl)+ G(h - hl). (9.30)
Blood insoluble/soluble 11--22
Hydroxide gels 6-10 Combining Eqs. (9.22) and (9.23) relates the saturation enthalpy
Riboflavin 15
and temperature,
Tannin 16-20

4. RESINS h, = h + (k,/kh)(T - T,). (9.31)

Acrylics 10-11
Formaldehyde resin 18-28 In Figure 9.15(c), Eq. (9.31) is represented by the line sloping
Polystyrene 12-15 upwards to the left. The few data that apparently exist suggest that
the coefficient ratio is a comparatively large number. In the absence
5. CERAVICS of information to the contrary, the ratio commonly is taken infinite,
Alumina 11L15 which leads to the conclusion that the liquid film resistance is
Ceramic colours 10 negligible and that the interface is at the bulk temperature of the
water. For a given value of T, therefore, the value of h, in Eq.
(Bowen Engineering Inc.). (9.25) is found from the equilibrium relation (hs, T,) of water and
the corresponding value of h from the balance Eq. (9.30). When the
coefficient ratio is finite, a more involved approach is needed to find
the integrand which will be described.
The equilibrium relation between T, and h, is represented on
(b) Performance of the Pilot Unit as a Function of Product the psychrometric charts Figures 9.1 and 9.2, but an analytical
Numbera representation also is convenient. From Section 9.1,
h, = 0.24T, + (18/29)(0.45T, + 11OO)[p,/(l -ps)], (9.32)
3200
3000 where the vapor pressure is represented by
2800
ps = exp[11.9176 - 7173.9/(T, + 389.5)]. (9.33)
2600

24W Over the limited ranges of temperature that normally prevail in

2200
cooling towers a quadratic fit to the data,
ETU LE.
EVAP'N
2wo
h, = a + bT, + CT:
1900 may be adequate. Then an analytical integration becomes possible
18W for the case of infinite k,/kh. This is done by Foust et al. (1980) for
example.
1700
The Cooling Tower Institute (1967) standardized their work in
terms of a Chebyshev numerical integration of Eq. (9.25). In this
100
method, integrands are evaluated at four temperatures in the
80
interval, namely,
60
LE, HR
EVAP" T, + 0.1(T2- q), corresponding integrand 11,
T, + 0.4(T2- T1),
40
corresponding integrand I,,
30 (9.34)
Tl - 0.4(T2- TI), corresponding integrand 13,
20
4 5 6 7 8 10 12 14 16 18 20 22 24 26 28 7'' - 0.1(T, - Tl), corresponding integrand 14.
PRODUCTNUMBER (DRYING EFFECTIVENESS)
Then the integral is
"Example: For a material with product number= 10 and air inlet
temperature of 500°F. the evaporation rate is 53Ib/hr, input Btu/lb
evaporated = 1930, and the air outlet temperature is 180°F. /T:&=0.25(Tz- Tl)(ll+Z2+Z3+14). (9.35)
(Bowen Engineering).
 9.11. THEORY OF AIR-WATER INTERACTION IN PACKED TOWERS 279

EXAMPLE9.10 Enthalpy loss of air is

S i g a Spray Dryer on the Basis of Pilot Plant Data
Feed to a spray dryer contains 20% solids and is to be dried to 5% 0.1(69.8 - 28.0) = 4.2 Btu/lb.
moisture at the rate of 5001b/hr of product. Pilot plant data show
that a residence time of 6sec is needed with inlet air of 230"F, Exit enthalpy of air is
H=0.008Ib/lb, and exit at 100°F. Ambient air is at 70°F and is
heated with steam. Enthalpy loss to the surroundings is 10% of the h=69.8-4.2=65.6.
heat load on the steam heater. The vessel is to have a 60" cone. Air
rate and vessel dimensions will be found. At 100°F and this enthalpy, other properties are read off the
Enthalpy, humidity, and temperatures of the air are read off psychrometric chart as
the psychrometric chart and recorded on the sketch.
H = 0.0375 Ib/lb,
Dry 475pph
V = 14.9 cuft/lb.
Water 1900 pph
-
I Air rate is
I
1900 - 25
70 F A= = 63,559 lb/hr
230F H 0.0375 - 0.008
H = 0.008
h = 28.0 h = 69.8
V=17.6 1 I
3600 2

With a residence time of 6 sec, the dryer volume is

V, = 287(6) = 1721.4 cuft.

Make the straight side four times the diameter and the cone
100 F 60":

Dry 475pph 0.866~0~

1721.4 = 40(nD2/4) + ~- - 3.3683D3,
Water 23pph 12
Total 500 pph :. D = 8.0 ft.

When k,,,/k, +.m, evaluation of the integrands is straightforward. TOWER HEIGHT

When the coefficient ratio i s finite and known, this procedure may
be followed:
1. For each of the four values of T, find h from Eq. (9.30).
2. Eliminate h, between Eqs. (9.31) and (9.32) with the result
+ (k,/k,)(T - T,)
= 0.24T, + (18/29)(0.45T, + llOO)[P,/(l -p,.)]. (9.36)
3. Substitution of Eq. (9.33) into (9.36) will result in an equation
that has T, as the only unknown. This is solved for with the
Newton-Raphson method.
4. Substitution of this value of T, back into Eq. (9.31) will evaluate
hs.
5. The integrand l/(hs - h ) now may be evaluated at each
temperature and the integration performed with Eq. (9.35).
Example 9.11 employs this method for finding the number of
transfer units as a function of liquid to gas ratio, both with finite and
infinite values of k,/k,. The computer programs for the solution of
this example are short but highly desirable. Graphical methods have
been widely used and are described for example by Foust et al.
(1980).
The information that is ultimately needed about a cooling tower
design is the height of packing for a prescribed performance. This
equals the product of the number of transfer units by the height of
each one,
Z = (NTU)(HTU). (9.37)
Some HTU data for cooling tower packing have been published, for
example, those summarized on Figure 9.16. Other data appear in
the additional literature cited for this chapter. Several kinds of
tower fill made of redwood slats are illustrated in Figure 9.17. The
numbers N of such decks corresponding to particular NTUs and
(L/G)s are given by the equation
[(NTU) - 0.07](L/G)b
N= (9.38)
a
Values of a and b are given for each type of fill with Figure 9.17.
These data are stated to be for 120°F inlet water. Although the
authors state that corrections should be estimated for other
temperatures, they do not indicate how this is to be done. For
example, with deck type C, N T U = 2 and L / G = 1 . 2 : N =
(2 - 0.07)(1.2)0.m/0.092 = 23.4 decks, or a total of 31.2 ft since the
deck spacing is 16in. The data of Figure 9.16 are used in Example
9.11.
 280 DRYERS AND COOLING TOWERS

(a)

Equilibrium p

c
'
c
c-

balance
I h = f (T)

Figure 9.15. Relations in a packed continuous flow air-water contactor. (a) Sketch of the tower
with differential zone over which the enthalpy and material balances are made. (b) Showing
equilibrium and operating lines from which the integrand l/(h3 - h) can be found as a function
of liquid temperature T. (c) Showing interfacial conditions as determined by the coefficient
ratio k,/k, ; when this value is large, interfacial and saturation temperatures are identical.

9.12. COOLING TOWERS

Cooling of water in process plants is accomplished most eco-

nomically on a large scale by contacting it with air in packed
towers. For reasons of economy, the tower fill is of a highly open
structure. Efficient ring and structured packings of the sort used for
distillation and other mass transfer processes are too expensive and
exert too high a power load on the fans. Standard cooling tower
practice allows a maximum of 2in. of water pressure drop of the
air. Water loadings range 500-2000 Ib/(hr)(sqft) or 1-4 gpm/sqft.
Gas loadings range 1300-1800 Ib/(hr)(sqft) or between 300 and
400ft/min. The liquid to gas ratio L / G normally is in the range
0.75-1.50 and the number of transfer units or the tower
characteristic, HTU = k,Z/L, vary from 0.5 to 2.50.
The most common fill is of wooden slats of rectangular or
01 I 10
triangular cross section arranged as in Figure 9.17. Corrugated
CIL sheets of asbestos-concrete have some application and also PVC
Figure 9.16. Data of heights of transfer units of packings construction unless the temperatures are above 160°F.
characterized by the specific surface ad (sqft/cuft). The ordinate is Fan power consumption is the major operating cost and can be
G/k, = Z / J dh/(h, - h), which is related to the form of NTU used counterbalanced in part by greater investment in natural draft
in this chapter by construction. In the majority of process applications, fan-operated
towers are preferred. Very large installations such as those in power
plants employ chimney assisted natural draft installations. A limited
use of atmospheric towers is made in areas where power costs are
The equation of the London line is equivalent to especially high.
The main types of cooling towers are represented on Figure
HTU = 5.51(L/G)0.59. 9.18. Their chief characteristics and some pros and cons will be
(Sherwood et ai., 1975). discussed in order.
 9.12. COOLING TOWERS 281

EXAMPLE 9.11 or
Sizing of a Cooling Tower: Number of Transfer Units and
Height of Packing HTU = 5.51(L/G)0.59.
Water is to be cooled from 110 to 75°F by contact with air that
enters countercurrently at 90°F with a dewpoint of 60°F. The data of Tower height:
London et al. (1940) of Figure 9.16 for height of transfer unit are
applicable. Calculations will be made for two values of the Z = (HTU)(NTU).
coefficient ratio k,/k,, namely, 25 and m Btu/("F) (Ib dry air), of
Eq. (9.31). The effect of the ratio of liquid to gas rates, L/G, will For several values of L/G:
be explored.
Ll G 0.6 1 1.4 1.7
HTU (ft) 4.08 5.51 6.72 7.54
T, = 110 F
Evaluation of interfacial temp and the NTU for L/G = 1 with
k,/k, = 25:

T h T. l/(k-h)
78.5 30.5 78.099 0.0864
89 41 88.517 0.0709
96 48 95.400 0.0575
106.5 58.5 105.581 0.0385

0.2533

:. NTU = (110 - 75(0.2533)/4= 2.217.

H = 0.011
For other values of L/G:
h =27

The maximum allowable L/G corresponds to equilibrium 1l(k- h )

between exit air and entering water at 110. The saturation enthalpy
T h L/G=0.6 1 1.4 1.7
at 110°F is 92, so that Eq. (9.30) becomes
78.5 30.5 0.0751 0.0864 0.0943 0.1043
89 41 0.0518 0.0709 0.1167 0.2200
mru 110-75 96 48 0.0398 0.0575 0.1089 0.3120
106.5 58.5 0.0265 0.0385 0.0724 0.1987
----
The several trials will be made at L/G = (0.6, 1.0, 1.4, 1.7). 0.1933 0.2533 0.3923 0.8350
The applicable equations with numerical substitutions are listed NTU -+ 1.691 2.217 3.433 7.306
here and incorporated in the computer program for solution of this
problem [Eqs . (9.30)- (9.33)]: With k,/kh-+m:

h = 27 + (L/G)( T - 75),
h, = h + 25(T - 75),
T h L/G=O.6 1 1.4 1.7
h, =0.24T + (18/29)(0.457 + 1100)</(1 -e),
< = exp[11.9176- 7173.9/(T, + 389.5))
78.5
89
30.5
41
0.0725
0.0494
0.0807
0.0683
0.90
0.1 107
0.1006
0.2070
96 48 0.0376 0.0549 0.1 020 0.2854
When k,/k,+m, T, in Eq. (9.33) is replaced by T.
106.5 106.5 0.0248 0.0361 0.0663 0.1778
The four temperatures at which the integrands are evaluated
for the Chebyshev integration are found with Eq. (9.34) and 0.1844 0.2400 0.3700 0.7708
tabulated in the calculation summary following. NTU-, 1.613 2.100 3.238 6.745
Equations (9.30) and (9.31) are solved simultaneously for h and Z+ 6.58 11.57 21.76 50.86
h, with the aid of the Newton-Raphson method as used in the
computer program; the integrands are evaluated and the integration
are completed with Eq. (9.35).
The number of transfer units is sensitive to the value of L / G ,
but the effect of k,/k, is more modest, at least over the high range
used; data for this ratio do not appear to be prominently recorded.
Figure 9.16 shows a wide range of heights of transfer units for the
different kinds of packings, here characterized by the surface ad
(sqft/cuft) and substantial variation with L/G. The last line of the
calculation summary shows variation of the tower height with L/G.
Data of London et al. (1940) of Figure 9.16:

(G/L)(HTU) = 5.51(G/L)0.41
282 DRYERS AND COOLING TOWERS

a. Atmospheric towers are effective when prevailing wind velocities

DECKS A B E DECKS C 8 D
are 5 miles/hr or more. For access to the wind they are narrow
k?.l/8"~3/8"1
?/E"
but long, lengths of 2000 ft having been constructed. Water drift
losses are relatively large. The savings because of elimination of

ilB''u''~ tall chimney or fan power is counterbalanced by increased size

because of less efficient cross flow and variations in wind
velocity.
b. Chimney assisted natural draft towers also eliminate fans. Most
VERTICAL SPACING A:9",8: 12" VERTICAL SPACING C:l6", D124"
DECK E DECK F of the structure is the chimney, the fill occupying only 10-12%
of the tower height at the bottom. The temperature and
humidity of the air increase as the air flows upward so that its
buoyancy increases and results in rapid movement through the
chimney. Smaller units are made as circular cylinders since these
can be built rapidly. The hyperboloidal shape has greater
3-1/4"

VERTICAL SPACING 2 4 "

f
VERTICAL SPACING 24"
strength for a given wall thickness. In towers as large as
250ftdia and 450ft high, wall thicknesses of 5-8in. of rein-
DECK H forced concrete are adequate. The enlarged cross section at

*
the top converts some kinetic energy into pressure energy which
assists in dispelling the exit humid air into the atmosphere.
The ratio of base diameter to height is 0.75-0.85, the ratio
of throat and base diameters is 0.55-0.65, and the ratio of
vertical depth of air opening to base diameter is 0.1-0.12. Air
VERTICAL SPACINO 24" IVERTICAL
I
SPACING 2 4 " velocity through the tower is 3-6ft/sec, water flow rates range
DECK I DECK J from 600 to 1800Ib/(hr)(sqft). Two towers each 375ft high are
I/2 "I I " ~7/8"17/8~ able to service a 500 MW power plant. Natural draft towers are
uneconomical below heights of 70 ft. The upper limit is imposed
principally by environmental visual considerations; towers 500 ft
high are in existence. A cost comparison is made with item d.
c. Hyperbolic fan assisted towers can have as much as three times
V E R T I C A L S P A C I N GI - l24"
/e" the capacity of the same size natural draft towers. The fans
VERTICAL SPACINQ 2 4 " provide greater control than the natural draft systems; for
example, they may be turned on only at peak loads. Rules of
Factors in Eq. 9.38 for the Number of Decks thumb cited by Cheremisinoff and Cheremisinoff (1981) for
Deck Type a b
relative sizing is that fan assisted hyperbolic towers may have
diameters 2/3 and heights 1/2 those of purely natural draft designs.
A 0.060 0.62 d. Countercurrent-induced draft construction is the most widely
B 0.070 0.62 used type in process industries. Mechanical draft is capable of a
C 0.092 0.60
greater degree of control than natural draft and such towers are
D 0.119 0.58
able in some cases to cool the water within 2°F of the wet bulb
E 0.110 0.46
F 0.100 0.51 temperature of the air. The elevated fan location introduces
G 0.104 0.57 some structural and noise problems. The flow of air is quite
H 0.127 0.47 uniform across the cross section and its discharge is positive and
I 0.135 0.57 at high velocity so that there is little backflow of humid air into
J 0.103 0.54 the tower. A cost comparison (dated 1978) with hyperbolic
Figure 9.17. Kinds of fill made of redwood slats for cooling towers, towers is made by Singham (1983, Sec. 3.12.4.1). The case is for
and factors for determining the required number of decks with inlet a water rate of 6.1 m3/sec, cooling range of 8.S'C, approach of
water at 120°F (Cherembinof and Cheremisinof, 1981). lWC, and wet bulb of 17°C. The cost of the natural draft tower
 9.12. COOLING TOWERS 283
AIR OUTLET

t
I

(a)

Figure 9.18. Main types of cooling towers. (a) Atmospheric, dependent on wind velocity. (b) Hyperbolic stack natural draft. (c) Hyperbolic
assisted with forced draft fans. (d) Counterflow-induced draft. (e) Crossflow-induced draft. ( f ) Forced draft. ( 9 ) Induced draft with surface
precooler for very hot water; also called wet/dry tower. [ ( b ) - ( e )from Cherernisinof and Cheremisiinof, 1981).
- 95 - I I I I I

- ~-
U
90% WATER FLOW ,

60 65 70 75 BO 60 65 70 75 80
AIR WET-BULB TEMPERATURE (OF) AIR WET -BULB TEMPERATURE (OF) AIR WET-BULB TEMPERATURE ( O F 1

F i r e 9.19. Typical cooling tower performance curves (Cheremisinoff and Cheremisinoff, 1981).

TABLE 9.20. Selected Data Required with Bids of Cooling Towers

A. Cooling Tower 3. Type and manufacturer

1. Number of cells 4. Full load speed, rpm
2. Cell dimensions, ft, in. 5. Frame size
3. Tower length, ft, in. 6. Full load current, amps
4. Tower width, ft. in. 7. Locked rotor current amps
5. Tower height, ft, in. 8. Weight, Ib
6. Casing, material and dimensions
7. Structure, material and dimensions
8. Fill decks, material and dimensions H. Distribution System
9. Partitions and baffles, materials and 1. Number and size of inlet flanges
dimensions 2. Height of water inlet above curb, ft, in.
10. Drift eliminators, material and 3. Header material
dimensions 4. Lateral material
11. Fan stacks, material and dimensions 5. Nozzle, or downspout material
12. Fan deck, material and dimensions
13. Louvers, material and dimensions
14. Board feet of fill J. Design Performance
15. Board feet total tower 1. Pumping head from top of basin curb, ft
16. Height of fan stacks, ft, in. 2. Spray loss, max %
17. Post extension below curb, ft, in. 3. Evaporation loss, max %
18. Total shipping weight, Ib 4. Fill wetted surface, ftz
19. Total operating weight, Ib 5. Total wetted surface, ft2
6. Effective splash surface, ft2
B. Fans 7. Effective cooling volume, ft3 (from elirni-
1. Number of units nators to water level)
2. Type and manufacturer 8. Air volume per fan, cfm
3. Diameter, ft, in. 9. Static pressure, inches of water
4. Number of blades per fan 10. Output horsepower/motor/(turbine)
5. Blade material 11. Tower loading, gpm/ft2
6. Hub material
7. rpm
8. Tip speed, fpm K. Drawings and Performance Curves
9. Mechanical efficiency, % 1. Tower outline elevation
10. Static efficiency, % 2. Foundation outline
11. Weight, Ib 3. Fill rack details
4. Drift eliminator details
C. Motors 5. Tower sheeting arrangement
1. Number of units 6. A series of guaranteed performance curves within limits of CTI
2. Size, HP Test Procedure ATP-105, latest revision

(Excerpted from Cheremisinoff and Cheremisinoff, 1981).


was 1.2 M pounds and that of the mechanical draft was 0.75 M
pounds, but the fan power was 775kW. The opinion was
expressed that mechanical draft towers are more economical at
water rates below 1.25 m3/sec (19,800 gpm).
Crossflow induced draft offer less resistance to air flow and can
operate at higher velocities, which means that less power and
smaller cell sizes are needed than for counterflows. The shorter
travel path of the air makes them less efficient thermally. The
cross flow towers are made wider and less high, consequently
with some saving in water pumping cost.
Forced draft towers locate the fans near ground level which
requires simpler support structures and possibly lower noise
levels. A large space must be provided at the bottom as air inlet.
Air distribution is poor because it must make a 90" turn. The
humid air is discharged at low velocity from the top of the tower
and tends to return to the tower, but at the same time the drift
loss of water is less. The pressure drop is on the discharge side of
the fan which is less power-demanding than that on the intake
side of induced draft towers.
Wet-dry towers employ heat transfer surface as well as direct
contact between water and air. Air coolers by themselves are
used widely for removal of sensible heat from cooling water on a
comparatively small scale when cooling tower capacity is limited.
Since dry towers cost about twice as much as wet ones,
combinations of wet and dry sometimes are applied, particularly
when the water temperatures are high, of the order of 160°F, so
that evaporation losses are prohibitive and the plumes are
environmentally undesirable. The warm water flows first through
tubes across which air is passed and then enters a conventional
REFERENCES
Drying
1. W.L. Badger and J.T. Banchero, Introduction to Chemical Engineering,
McGraw-Hill, New York, 1955.
2. C.W. Hall, Dictionary of Drying, Dekker, New York, 1979.
3. R.B. Keey, Drying Principles and Practice, Pergamon, New York, 1972.
4. R.B. Keey, Introduction to Industrial Drying Operations, Pergamon,
New York, 1978.
5 . K. Kroll, Trockner und Trocknungsuerfahren, Springer-Verlag, Berlin,
1978.
6. P.Y. McCormick, Drying, in Encyclopedia of Chemical Technology,
Wiley, New York, 1979, Vol. 8, pp. 75-113.
7 . K. Masters, Spray Drying, George Godwin, London, 1976.
8. A S . Mujumdar (Ed.), Advances in Drying, Hemisphere, New York,
1980-1984, 3 vols.
9. G. Nonhebel and A.A.H. Moss, Drying of Soli& in the Chemical
Indrrstry, Butterworths, London, 1971.
10. R.E. Peck, Drying solids, in Encyclopedia of Chemical Processing and
Design, Dekker, New York, 1983, Vol. 17, pp. 1-29.
11. E.U. Schliinder, Dryers, in Heat Exchanger Design Handbook,
Hemisphere, New York, 1983, Sec. 3.13.
U.G.A. Schurr, Solids drying, in Chemical Engineers Handbook,
McGraw-Hill, New York, 1984, pp. 20.4-20.8.
l3. T.H. Wentz and J.R. Thygeson, Drying of wet solids, in Handbook of
Separation Techniques for Chemical Engineers, (Schweitzer, Ed.),
McGraw-Hill, New York, 1979.
REFERENCES 285
packed section where it is cooled further by direct contact with
air. Separate dampers for air to the dry and wet sections can
throw greater load on the wet section in summer months.
WATER FACTORS
Evaporation losses are about 1%of the circulation for every 10°F of
cooling range. Windage or drift losses are 0.3-1.0% for natural
draft towers and 0.1-0.3% for mechanical draft. Usually the salt
content of the circulating water is limited to 3-7 times that of the
makeup. Blowdown of 2.5-3% of the circulation accordingly is
needed to maintain the limiting salt concentration.
TESTING AND ACCEPTANCE
At the time of completion of an installation, the water and air
conditions and the loads may not be exactly the same as those of the
design specification. Acceptance tests performed then must be
analyzed to determine if the performance is equivalent to that under
the design specifications. Such tests usually are performed in
accordance with recommendations of the Cooling Tower Institute.
The supplier generally provides a set of performance curves
covering a modest range of variation from the design condition, of
which Figure 9.19 is a sample. Some of the data commonly required
with bids of cooling tower equipment are listed in Table 9.20, which
is excerpted from a 10-page example of a cooling tower requisition
by Cheremisinoff and Cheremisinoff (1981).
14. A. Williams-Gardner, Industrial Drying, Leonard Hill, Glasgow, 1971
Cooling Towers
1. N.P. Cheremisinoff and P.N. Cheremisinoff, Cooling Towers: Selection,
Design and Practice, Ann Arbor Science, Ann Arbor, MI, 1981.
2. Cooling Tower Institute, Performance Curves, (JII, Spring, TX,1967.
3. A.S. Foust et al., Principles of Unit Operations, Wiley, New York, 1980.
4. D.Q. Kern, Process Heat Transfer, McGraw-Hill, New York, 1950.
5. T.K. Sherwood, R.L. Pigford, and C.R. Wilke, Mars Transfer,
McGraw-Hill, New York, 1975.
6. J.R. Singham, Cooling towers, in Heat Exchanger Design Handbook,
Hemisphere, New York, 1983, Sec. 3.12.
Data on Performance of Cooling Tower Packing
1. Hayashi, Hirai, and Okubo, Heat Transfer Jpn. Res. 2(2) 1-6 (1973).
2. Kelly and Swenson, Chem. Eng. Prog. 52, 263 (1956), cited in Figure
9.16.
3. Lichtenstein, Trans. ASME 66, 779 (1943), cited in Figure 9.16.
4. London, Mason, and Boelter, Trans. ASME 62, 41 (1940), cited in
Figure 9.16.
5. Lowe and Christie, Proceedings, International Heat Transfer Conference,
Boulder, CO, 1961, Part V, pp. 933-950.
6. Simpson and Sherwood, Refrig. Eng. 52,535 (1946), cited in Figure 9.16.
7 . Tezuka, Heat Transfer Ipn. Res. 2(3), 4-52 (1973).