Chapter 18

Drying of Solids

§18.0 INSTRUCTIONAL OBJECTIVES

After completing this chapter, you should be able to:
Describe two common modes of drying.
Discuss industrial drying equipment.
Use a psychrometric chart to determine drying temperature.
Differentiate between the adiabatic-saturation and wet-bulb temperatures.
Explain equilibrium-moisture content of solids.
Explain types of moisture content used in making dryer calculations.
Describe the four different periods in direct-heat drying.
Calculate drying rates for different periods.
Apply models for a few common types of dryers.

D rying is the removal of moisture (either water or other vola-
tile compounds) from solids, solutions, slurries, and pastes to
give solid products. In the feed to a dryer, moisture may be:
embedded in a wet solid, a liquid on a solid surface, or a solu-
tion in which a solid is dissolved. The term drying also de-
scribes a gas mixture in which a condensable vapor is removed
from a non-condensable gas by cooling, as discussed in Chap-
ter 4, and the removal of moisture from a liquid or gas by sorp-
tion, as discussed in Chapters 6 and 16. This chapter deals only
with drying operations that produce solid products.
Drying is widely used to remove moisture from: (1) crystal-
line particles of inorganic salts and organic compounds
to produce a free-flowing product; (2) biological materials, in-
cluding foods, to prevent spoilage and decay from micro-
organisms that cannot live without water; (3) pharmaceuticals;
(4) detergents; (5) lumber, paper, and fiber products; (6)
dyestuffs; (7) solid catalysts; (8) milk; and (9) films and coatings,
and (10) products where high water content entails excessive
transportation and distribution costs. Not all drying processes
have been successful; the beer industry, for decades, has been
trying to market dehydrated beer with no success whatsoever.
Drying can be expensive, especially when large amounts
of water, with its high heat of vaporization, must be evapo-
rated. Water and energy conservation measures, and advan-
ces in equipment design, have broadened the use of pre-feed
dewatering operations by mechanical means such as expres-
sion; gravity, vacuum, or pressure filtration; settling; and cen-
trifugation, which also diminish the length of drying cycles.
Because drying involves vaporization of moisture, heat must
be transferred to the material being dried. The common modes
of heat transfer are: (1) convection from a hot gas in contact
with the material; (2) conduction from a hot, solid surface in
contact with the material; (3) radiation from a hot gas or surface;
and (4) heat generation within the material by dielectric, radio
frequency, or microwave heating. These different modes can
sometimes be used symbiotically, depending on whether the
moisture to be removed is on the surface or inside the solid.
Of importance is the temperature at which the moisture
evaporates. When convection from a hot gas is employed and
the moisture is on the surface or rapidly migrates to the sur-
face from the interior of the solid, the rate of evaporation is
independent of the properties of the solid and is governed by
the rate of convective heat transfer from the gas to the sur-
face. Then, the evaporating surface is at the wet-bulb temper-
ature of the gas if the dryer operates adiabatically.
If the convective heat transfer is supplemented by radia-
tion, the temperature of the evaporating surface is higher
than the wet-bulb temperature. In the absence of contact with
a convective-heating gas, as in the latter three modes, and
when a sweep gas is not present, such that the dryer operates
nonadiabatically, the evaporating moisture is at its boiling-
point temperature at the pressure in the dryer. If the moisture
contains dissolved, nonvolatile substances, the boiling-point
temperature will be elevated.
Industrial Example
The continuous production of 69,530 lb/day of MgSO4 7H2O
crystalline solids containing 0.015 lb H2O/lb dry solid is an
example of an industrial drying operation. The feed to the
dryer in Figure 18.1 consists of a filter cake from a rotary-

726

Air out
155°F
0.0204 lb H2O/lb dry air
Direct-heat ro
tary dryer
Filter cake
85°F
20.5 wt% moisture
(wet basis)
5-ft diameter × 30-ft length
4-rpm rotation
heat duty = 865,000 Btu/h
694 lb/h H2O evaporated
drum vacuum filter (see §19.2.7). The cake is at 85 F and
contains 20.5 wt% moisture on a wet basis. Because the feed
crystals are relatively coarse, free flowing, and nonsticking, a
direct-heat rotary dryer consisting of a slightly inclined,
rotating, cylindrical shell is used. The filter-cake feed enters
the high end of the dryer from an inclined, vibrated chute.
Heated air at 250 F and atmospheric pressure, with an abso-
lute humidity of 0.002 lb H2O/lb dry air, enters the other end
at a flow rate of 37,770 lb/h. To obtain good contact between
the wet crystals and hot air, the dryer is provided with inter-
nal, longitudinal flights that extend the entire shell length. As
the shell rotates, the flights lift the solids until they reach
their angle of repose and then shower down through the hot
air in countercurrent flow to the direction of net movement of
the solids. The dry solids discharge at 113 F through a rotat-
ing valve into a screw conveyor. The air, which has been
cooled to 155 F and humidified to 0.0204 lb H2O/lb dry air
by contact with the wet solids, exits at the other end through
a fan to pollution-control units.
The hot air causes evaporation of 694 lb/h of water,
mostly at a temperature of 94.5 F, which is the average of
the entering and exiting gas wet-bulb temperatures of 95.5
and 93.5 F, respectively. In addition, the hot air must heat the
solids from 85 F to 113 F and the evaporated moisture to
155 F. The total rate of convective heat transfer, Q, from the
gas to the solids is 865,000 Btu/h. This ignores heat loss from
the dryer shell to the surroundings and thermal radiation to
the solids from the hot gas or the inside shell surface.
Of the total heat load, approximately 83% is required to
evaporate moisture, with the balance supplying sensible heat.
Therefore, a reasonably accurate log mean temperature-
driving force is based on the assumption of a constant tem-
perature at the gas–wet solids interface equal to the average
air wet-bulb temperature of 94.5 F:
ð250 94:5Þ ð155 94:5Þ
DT LM ¼ ¼ 100:6 F
250 94:5
ln
155 94:5
For a direct-heat, rotary dryer, convective heat transfer is
characterized by an overall volumetric heat-transfer co-
efficient, Ua, which for this example is 14.6 Btu/h-ft3 of
dryer volume- F. The required cylindrical shell volume, V,
from Q ¼ UaVDT LM , is 590 ft3 and the dryer diameter is
5 ft, which gives an entering superficial hot-air velocity of
§18.1 Drying Equipment 727
Air in
250°F, 1 atm
37,770 lb/h
0.002 lb H2O/lb dry air
69,530 lb/day
magnesium sulfate
heptahydrate crystals
Figure 18.1 Process for drying
113°F
1.5 wt% moisture magnesium–sulfate–heptahydrate
(dry basis) filter cake.
9.56 ft/s, which is sufficiently low to prevent entrainment of
solid particles in the air. The cylindrical shell is 30 ft long
and rotates at 4 rpm. While moving through the dryer, the
bulk solids, with a bulk density of 62 lb dry solids/ft3, occupy
8 vol% of the dryer, and have a residence time of one hour.
_________________________________________________
§18.1 DRYING EQUIPMENT
Material sent to drying equipment includes granular solids,
pastes, slabs, films, slurries, fabrics, and liquids. Accord-
ingly, different types of feed- and product-specific dryers
have been developed.
§18.1.1 Classification of Dryers
Dryers can be classified in a number of ways; perhaps most
importantly is the mode of operation, batch or continuous.
Batch operation is generally indicated when the production
rate is less than 500 lb/h of dried solid, while continuous
operation is preferred for a production rate of more than
2,000 lb/h. In the example above, the production rate is
2,900 lb/h and continuous drying was selected.
A second classification method is the mode of heat trans-
fer to evaporate moisture. As mentioned, direct-heat (also
called convective or adiabatic) dryers contact material with a
hot gas, which not only provides the energy to heat the mate-
rial and evaporate the moisture, but also sweeps away the
moisture. When the continuous mode of operation is used,
the hot gas can flow countercurrently, cocurrently, or in
crossflow to the material being dried. Countercurrent flow is
the most efficient, but cocurrent flow may be required if the
material being dried is temperature-sensitive.
Indirect-heat (also called nonadiabatic) dryers provide
heat to the material by conduction and/or radiation from a
hot surface. Energy may also be generated within the mate-
rial by dielectric, radio frequency, or microwave heating.
Indirect-heat dryers may be operated under vacuum to reduce
the temperature at which the moisture is evaporated. A sweep
gas is not necessary, but can be provided to help remove
moisture. Capital costs for direct-heat dyers are higher, but
indirect-heat dryers are more expensive to operate and are
used only when the material is either temperature-sensitive
or subject to crystal breakage and dust or fines formation.
728 Chapter 18 Drying of Solids

A third method for classifying dryers is the degree to
which the material is agitated. In some dryers, the feed is sta-
tionary while being processed. At the opposite extreme is the
fluidized-bed dryer, in which agitation increases the rate of
heat transfer but, if too severe, can cause crystal breakage
and dust formation. Agitation may be necessary if the mate-
rial is sticky.
The more widely used commercial dryers are described
here. A more complete coverage is given in the Handbook of
Industrial Drying [1]. Extensive performance data for many
types of dryers are given in Perry’s Chemical Engineers’
Handbook [2] and by Walas [3]. Batch dryers are discussed
first, followed by continuous dryers, and then other dryers
that use special means for evaporating moisture.
§18.1.2 Batch Dryers
Equipment for drying batches includes: (1) tray (also called
cabinet, compartment, or shelf) dryers; and (2) agitated
dryers. Together, these two types cover many of the modes of
heat transfer and agitation discussed above.
Tray Dryers
The oldest and simplest batch dryer is the tray dryer, which is
shown schematically in Figure 18.2 and is useful when low
production rates of multiple products are involved and when
drying times vary from hours to days. The material to be
dried is loaded to a depth of typically 0.5–4 inches in remov-
able trays that may measure 30 30 3 inches and are
Adjustable
Trays louvers
Air in Air out
Fan
Heater
Screen
(a) Cross-circulation
stacked by a forklift on shelves about 3 inches apart in a cabi-
net. If the wet solids are granular or shaped into briquettes,
noodles, or pellets with appreciable voids, the tray bottom
can be perforated so that heating gas can be passed down
through the material (through-circulation) as shown in Fig-
ure 18.2b. Otherwise, the tray bottom is solid and the hot gas
is passed at velocities of 3–30 ft/s over the open tray surface
(cross-circulation), as in Figure 18.2a.
Although fresh hot gas might be used for each pass
through the dryer, it is more economical to recirculate the
gas, providing venting and makeup gas at rates of 5–50% of
the circulation rate to maintain the humidity at an acceptable
level. Gas is heated with an annular, finned-tube heat
exchanger by steam condensing inside the tubes. If the mois-
ture being evaporated is water, steam requirements can range
from 1.5 to 7.5 lb steam/lb water evaporated. It is important
to baffle tray dryers to promote uniform distribution of hot
gas to achieve uniform drying.
Tray dryers are available for vacuum operation and with
indirect heating. In one configuration, the trays are placed on
hollow shelves that carry condensing steam and act as heat
exchangers. Heat is transferred by conduction to a tray from
the top of the shelf supporting it and by radiation from the
bottom of the shelf directly above the tray. Typical perform-
ance data for direct-heat, crossflow-circulation tray dryers are
given in Table 18.1.
Agitated Dryers
As discussed by van’t Land [4] and Uhl and Root [5], indirect
heat with agitation and, perhaps, under vacuum, is desirable
for batch drying when any of the following conditions exist:
(1) material oxidizes or becomes explosive or dusty during
drying; (2) moisture is valuable, toxic, flammable, or explo-
sive; (3) material tends to agglomerate or set up if not agi-
tated; and (4) maximum product temperature is less than
about 30 C. Heat-transfer rates are controlled mostly by con-
tact resistance at the inner wall of the jacketed vessel and by
conduction into the material being dried. A wide variety of
heating fluids can be used, including hot liquids, steam, Dow-
therm, hot air, combustion gases, and molten salt.
When only Condition 3 applies, the atmospheric, agitated-
pan dryer shown in Figure 18.3a is useful, particularly when

Table 18.1 Performance Data for Direct-Heat, Crossflow-

Air Circulation Tray Dryers

Aspirin-Base
Material Granules Chalk Filter Cake

Number of trays 20 72 80
Trays Area/tray, ft2 3.5 15.7 3.5
Total loading, lb wet 56 1,800 2,800
Depth of loading, inches 0.5 2.0 1.0
Fan % Initial moisture 15 46 70
% Final moisture 0.5 2 1
Maximum air temp., F 122 180 200
(b) Through-circulation
Drying time, h 14 4.5 45
Figure 18.2 Tray dryers.

Steam
supply
Bearings
Paddle
Vessel
(a) Atmospheric pan dryer
(b) Rotating, double-cone vacuum dryer
Shaft
drive
Shaft
oscillator
Steam Valves
jacket
Valves
Product
(c) Paddle-agitated cylinder dryer
Figure 18.3 Agitated dryers.
[From Perry’s Chemical Engineers’ Handbook, 6th ed., R.H. Perry, D.W. Green, and J.O. Maloney, Eds., McGraw-Hill, New York (1984) with permission.]
the feed is a liquid, slurry, or paste. This dryer consists of a
shallow (2–3-ft high), jacketed, flat-bottomed vessel,
equipped with a paddle agitator that rotates at 2–20 rpm and
scrapes the inner wall to help prevent cake buildup. Units
range in size from 3 to 10 ft in diameter, with a capacity of
up to 1,000 gallons and from 15 to 300 ft 2 of heat-transfer
surface. When using steam in the jacket, overall heat-transfer
coefficients vary from 5 to 75 Btu/h-ft 2- F. The material to be
dried occupies about 2/3 of the vessel volume. The degree of
agitation can be varied during the drying cycle. With a thin-
liquid feed, agitation may vary from very low initially to very
high if a sticky paste forms, followed by moderate agitation
when the granular solid product begins to form. Typically,
several hours are required for drying. Vacuum units are also
available.
When any or all of the above four conditions apply, but
only mild agitation is required, the jacketed, rotating, dou-
ble-cone (also called tumbler) vacuum dryer, shown schemat-
ically in Figure 18.3b, can be used. V-shaped tumblers are
also available. The conical shape facilitates discharge of
dried product, but, except for the tumbling, no means is
§18.1 Drying Equipment 729
Hatch
Steam jacket
Vacuum ducts
Drive
Vapors Vapor
filter
Feed
Paddles
Shaft with
paddles
Steam
jacket
provided to prevent cake buildup on the inner walls. Double-
cone volumes range from 0.13 to 16 m3, with heat-transfer
surface areas of 1 to 56 m2. Additional heat-transfer surface
can be provided by internal tubes or plates. Up to 70% of the
volume can be occupied by feed. A typical evaporation rate
when operating at 10 torr, with heating steam at 2 atm, is 1
lb/h-ft2 of heat-transfer surface.
A more widely used agitated dryer, applicable when any
or all of the above four conditions are relevant, is the ribbon-
or paddle-agitated, horizontal-cylinder dryer, shown in the
paddle form in Figure 18.3c. The cylinder is jacketed and sta-
tionary. The ribbons or paddles provide agitation and scrape
the inner walls to prevent solids buildup. As discussed by Uhl
and Root [5], cylinder dimensions range up to diameters of
6 ft and lengths up to 40 ft. The agitator can be rotated from
4 to 140 rpm, resulting in overall heat-transfer coefficients of
5 to 35 Btu/h-ft2- F. Typically from 20 to 70% of the cylinder
volume is filled with feed, and drying times vary from 4 to 16
hours. In more advanced versions, discussed by McCormick
[6], one or two parallel rotating shafts can be provided that
intermesh with stationary, lump-breaking bars to increase the
730 Chapter 18 Drying of Solids
range of application. The paddles can also be hollow to pro-
vide additional heat-transfer surface. This type of dryer can
also be operated in a continuous mode.
§18.1.3 Continuous Dryers
A wide variety of industrial drying equipment for continuous
operation is available. The following descriptions cover most
types, organized by the nature of the wet feed: (1) granular,
crystalline, and fibrous solids, cakes, extrusions, and pastes;
(2) liquids and slurries; and (3) sheets and films. In addition,
infrared, microwave, and freeze-drying are described.
Tunnel Dryers
The simplest, most widely applicable, and perhaps oldest
continuous dryers are the tunnel dryers, which are suitable
for any material that can be placed into trays and is not sub-
ject to dust formation. The trays are stacked onto wheeled
trucks, which are conveyed progressively in series through a
tunnel where the material in the trays is contacted by cross-
circulation of hot gases. As shown in Figure 18.4, the hot
gases can flow countercurrently, cocurrently, or in more com-
plex flow configurations to the movement of the trucks. As a
truck of dried material is removed from the discharge end of
the tunnel, a truck of wet material enters at the feed end. The
overall drying operation is not truly continuous because wet
material must be loaded into the trays and dried material
removed from the trays outside the tunnel, often with dump
truck devices. Tray spacings and dimensions, as well as hot-
gas velocities, are the same as for batch tray dryers. A typical
tunnel might be 100 ft long and house 15 trucks.
Belt or Band Dryers
A more continuous operation can be achieved by carrying the
solids as a layer on a belt conveyor, with hot gases passing
over the material. The endless belt is constructed of hinged,
Heater
Fresh
air
inlet
Wet
material
in Trucks
Exhaust-air stack
(a) Countercurrent flow
Blower
Wet
material
in
Trucks
(b) Cocurrent flow
slotted-metal plates, or, preferably a thin metal band, which
is ideal for slurries, pastes, and sticky materials. The bands
are up to 1.5 m wide 1 mm thick.
More common are screen or perforated-belt or band-
conveyor dryers, which, as shown in Figure 18.5a, use circu-
lation of heated gases upward and/or downward through a
moving, permeable, layered bed of wet material from 1 to 6
inches deep. As shown in Figure 18.5b, multiple sections,
each with a fan and set of gas-heating coils, can be arranged
in series to provide a dryer, with a single belt as long as 150 ft
with a 6-ft width, giving drying times up to 2 h, with a belt
speed of about 1 ft/minute. To be permeable, the wet material
must be granular. If it is not, the material can be preformed
by scoring, granulation, extrusion, pelletization, flaking, or
briquetting. Particle sizes usually range from 30 mesh to 2
inches. Hot-gas superficial velocities through the bed range
from 0.5 to 1.5 m/s, with maximum bed pressure drops of
50-mm of water. Heating gases are provided by heat transfer
from condensing steam in finned-tube heat exchangers at 50–
180 C, but temperatures up to 325 C are feasible. Continu-
ous, through-circulation conveyor dryers are used to remove
moisture from a variety of materials, some of which are listed
in Table 18.2, which includes, in parentheses, the method of
preforming, if necessary.
A perforated-band-conveyor dryer 50-ft long 75-inches
wide can produce 1,800 lb/h of calcium carbonate with a
moisture content of 0.005 lb H2O/lb carbonate, in a residence
time of 40 minutes, from 6-mm-diameter carbonate extru-
sions with a moisture content of 1.5 lb H2O/lb carbonate,
using air heated to 320 F by 160 psig steam and passing
through the bed of extrusions at a superficial velocity of 2.7
ft/s. Steam consumption is 1.75 lb/lb H2O evaporated.
Turbo-Tray Tower Dryers
When floor space is limited but headroom is available, the
turbo-tray or rotating-shelf dryer, shown in Figure 18.6, is a
good choice for rapid drying of free-flowing, nondusting
Blower
Dry
material
out
Heater
Fresh
air
inlet
Dry
material
out
Exhaust-air stack
Figure 18.4 Tunnel dryer.
 §18.1 Drying Equipment 731

(a) Single downflow section

Figure 18.5 Perforated-belt or

(b) Multiple sections band-conveyor dryer.

Table 18.2 Materials Dried in Through-Circulation Conveyor

Dryers Turbines (fans)
Feed
Heating elements
Aluminum hydrate (scored on filter)
Aluminum stearate (extruded)
Asbestos fiber 1
Breakfast food
Calcium carbonate (extruded)

Drying zones
Cellulose acetate (granulated)
Charcoal (briquetted) 2
Cornstarch
Cotton linters
Cryolite (granulated)
Dye intermediates (granulated) 3
Fluorspar
Gelatin (extruded)

Cooling zone
Kaolin (granulated)
Lead arsenate (granulated)
Circular
Lithopone (extruded)
shelves
Discharge
Magnesium carbonate (extruded)
Mercuric oxide (extruded)
(a) Turbo-tray tower dryer
Nickel hydroxide (extruded)
Polyacrylic nitrile (extruded)
Rayon staple and waste
Sawdust
Scoured wool
Turbo fan
Silica gel
Soap flakes Slots

Soda ash
Starch (scored on filter)
Sulfur (extruded)
Stationary
Synthetic rubber (briquetted) wiper Stationary
Pile of material leveler
Tapioca Material falling
to tray below from tray above
Titanium dioxide (extruded) (b) Detail of annular shelf
Zinc stearate (extruded)
Figure 18.6 Rotating-shelf dryer.
732 Chapter 18 Drying of Solids

granular solids. Annular shelves, mounted one above the
other, are slowly rotated at up to 1 rpm by a central shaft.
Wet feed enters through the roof onto the top shelf as it
rotates under the feed opening. At the end of one revolution,
a stationary wiper causes the material to fall through a radial
slot onto the shelf below, where it is spread into a pile of uni-
form thickness by a stationary leveler. This action is repeated
on each shelf until the dried material is discharged from the
bottom of the unit. Also mounted on the central shaft are fans
that provide cross-circulation of hot gases at velocities of 2 to
8 ft/s across the shelves, and heating elements located at the
unit’s outer periphery. The bottom shelves can be used as a
solids-cooling zone. Because solids are showered through
the hot gases and redistributed from shelf to shelf, drying
time is less than for cross-circulation, stationary-tray dryers.
Typical turbo-tray dryers are from 2 to 20 m in height and 2
to 11 m in diameter, with shelf areas to 1,675 m2. Overall
heat-transfer coefficients (based on shelf area) of 30–120
J/m 2-s-K have been observed, giving moisture-evaporation
rates comparable to those of through-circulation, belt-, or
band-conveyor dryers. Materials successfully handled in
turbo-tray dryers include calcium hypochlorite, urea, calcium
chloride, sodium chloride, antibiotics, antioxidants, and
water-soluble polymers. Capacities of up to 24,000 lb/h of
dried product are quoted.
Direct-Heat Rotary Dryers
A popular dryer for evaporating water from free-flowing
granular, crystalline, and flaked solids of relatively small
size, when breakage of solids can be tolerated, is the direct-
heat rotary dryer. As shown in Figure 18.7a, it consists of a
rotating, cylindrical shell that is slightly inclined from the
horizontal with a slope of less than 8 cm/m. Wet solids enter
through a chute at the high end and dry solids discharge from
the low end. Hot gases (heated air, flue gas, or superheated

A — Dryer shell
B — Shell-supporting rolls
C — Drive gear
A
D — Gas-discharge hood
G G E — Exhaust fan
F B F — Feed chute
Moist Feed G — Lifting flights
air A J — Air heater
outlet

D C J Air
B Steam inlet

E Steam
condensate
Dry solids discharge
(a) Rotary dryer

Radial 45° lip

flights flights
(b) Lifting flights

Exhaust-gas
outlet
Wet-feed
inlet

Hot-air
inlet

Product
Air flow-through outlet Figure 18.7 Direct-heat rotary dryer.
louvers and
[From W.L. McCabe, J.C. Smith, and P. Harriott, Unit
material
Operations of Chemical Engineering, 5th ed., McGraw-
Hill, New York (1993) with permission.] [From Perry’s
Chemical Engineers’ Handbook, 6th ed., R.H. Perry,
Hot-air chambers D.W. Green, and J.O. Maloney, Eds., McGraw-Hill,
(c) Roto-louvre dryer New York (1984) with permission.]
 §18.1 Drying Equipment 733

Table 18.3 Materials Dried in Direct-Heat Rotary Dryers
Ammonium nitrate prills Sand
Ammonium sulfate Sodium chloride
Blast furnace slag Sodium sulfate
Calcium carbonate Stone
Cast-iron borings Polystyrene
Cellulose acetate Sugar beet pulp
Copper Urea crystals
Fluorspar Urea prills
Illmenite ore Vinyl resins
Oxalic acid Zinc concentrate
steam) flow countercurrently to the solids, but cocurrent flow
can be employed for temperature-sensitive solids. With
cocurrent flow, the cylinder may not need to be inclined
because the gas will help move the solids. To enhance the gas-
to-solids heat transfer, longitudinal lifting flights—available in
several different designs, two of which are shown in Figure
18.7b—are mounted on the inside of the rotating shell, causing
the solids to be lifted, then showered through the hot gas during
each cylinder revolution. Typically the bulk solids occupy 8–
18% of the cylinder volume, with residence times from 5 min-
utes to 2 h. Resulting water-evaporation rates are 5–50 kg/h-m3
of dryer volume. The gas blower can be located to push or pull
the gas through the dryer, with the latter favored if the material
tends to form dust. Knockers, on the outside shell wall, can be
used to prevent solids from sticking to the inside shell wall.
Rotary dryers are available from 1 to 20 ft in diameter and 4–
150 ft long. Superficial-gas velocities, which may be limited
by dust entrainment, are 0.5–10 ft/s. The peripheral shell ve-
locity is typically 1 ft/s. A variety of materials, some of
which are listed in Table 18.3, are dried in direct-heat rotary
dryers. The detailed mechanical designs of rotary dryers are
industry specific in the sense that the standard designs are
modified to accommodate starch, sugar, salt, cement and
other products, each of which has unique surface and bulk
properties.
Roto-Louvre Dryers
A further improvement in the rate of heat transfer from hot
gas to solids in a rotating cylinder is the through-circulation
action achieved in the Roto-Louvre dryer in Figure 18.7c. A
double wall provides an annular passage for hot gas, which
passes through louvers and then through the rotating bed of
solids. Because gas pressure drop through the bed may be
significant, both inlet and outlet gas blowers are often pro-
vided to maintain an internal pressure close to atmospheric.
These dryers range from 3 to 12 ft in diameter and 9–36 ft
long, with water-evaporation rates reported as high as 12,300
lb/hr. They are useful for processing coarse, free-flowing,
dust-free solids.
Indirect-Heat, Steam-Tube Rotary Dryers
When materials are: (1) free flowing and granular, crystal-
line, or flaked; (2) wet with water or organic solvents; and/or
(3) subject to undesirable breakage, dust formation, or con-
tamination by air or flue gases, an indirect-heat, steam-tube
rotary dryer is often selected. A version of this dryer, shown
in Figure 18.8, consists of a rotating cylinder that houses two
concentric rows of longitudinal finned or unfinned tubes that
carry condensing steam and rotate with the cylinder. Wet sol-
ids are fed into one end of the cylinder through a chute or by
a screw conveyor. A gentle solids-lifting action is provided
by the tubes. Dried product discharges from the other end

tation
Ro

Dust drum
Section of "A-A" Section through steam manifold

Steam
Wet A manifold
material
fed in here Steam
neck

A Dried
material
discharge conveyor

Figure 18.8 Indirect-heat, steam-tube rotary dryer.

[From Perry’s Chemical Engineers’ Handbook, 6th ed., R.H. Perry, D.W. Green, and J.O. Maloney, Eds., McGraw-Hill, New York (1984) with permission.]
734 Chapter 18 Drying of Solids
Wet solids
Condensate
Figure 18.9 Screw-conveyor dryer.
after suitable contact with the hot-tube surfaces. The mois-
ture (water or solvent) evaporates at about the boiling tem-
perature, but can be swept out by a purge of inert gas. Steam
enters the tubes through a central revolving inlet manifold.
Condensate is discharged into a collection ring. With
unfinned tubes, overall heat-transfer coefficients based on the
surface area of the tubes range from 30 to 85 J/m2-s-K, when
solids occupy 10–20% of the dryer volume. Steam-tube
rotary dryers range in size from 3 to 8 ft in diameter by 15–80
ft long, with one or two rows of 14–90 tubes, 2.5–4.5 inches
in diameter. The largest-size dryers contain a single row of 90
tubes. Rotation rates are from 3 to 6 rpm. Materials success-
fully dried include inorganic crystals, silica, mica, flotation
concentrates, pigment filter cakes, precipitated calcium car-
bonate, distillers’ grains, brewers’ grains, citrus pulp, cellu-
lose acetate, starch, and high-moisture organic compounds.
Screw-Conveyor Dryers
Less popular than rotary dryers is the screw-conveyor dryer,
shown in Figure 18.9, which consists of a trough or cylinder
that carries a hollow screw, inside of which steam condenses
to provide heat for drying the material being conveyed. Addi-
tional heat transfer can be provided by jacketing the trough or
cylindrical shell. A wide range of materials can be dried,
including slurries, solutions, and solvent-laden solids. The
boiling moisture can be purged with a small amount of inert
gas. Standard conveyor dryers are as large as 3 ft in diameter
by 20 ft long. More drying time can be provided by arranging
a number of units in series, with one unit above another to
save floor space. The last unit can be a cooler. Overall heat-
transfer coefficients are comparable to, but less than, those
for indirect-heat, steam-tube rotary dryers. Major applica-
tions include removal of solvents from solids and drying of
fine and sticky materials.
Fluidized-Bed Dryers
Free-flowing, moist particles can be dried continuously with
a residence time of a few minutes by contact with hot gases
Steam
Condensate
Steam
Dry solids
in a fluidized-bed dryer, such as that shown in Figure 18.10a.
This dryer consists of a cylindrical or rectangular fluidizing
chamber to which wet particles are fed from a bin through a
star valve or by a screw conveyor, and fluidized by hot gases
blown through a heater and into a plenum chamber below the
bed, from where the particles pass into the fluidizing cham-
ber through a distributor plate, which must have a pressure
drop of from 15 to 50% of the static bed head. The hot gases
pass up through the bed, transferring heat for evaporation of
the moisture, and pass out the top of the fluidizing chamber
and through demisters and cyclones for dust removal. The
solids are circulated by the action of the hot gases in the bed
and by baffles, and sometimes mixers, but eventually pass out
of the chamber through an overflow duct, which also serves
to establish the height of the fluidized bed.
Fluidized-bed dryers have become very popular in
recent years because they: (1) have no moving parts;
(2) provide rapid heat and mass transfer between gas and
particles; (3) provide intensive mixing of the particles, lead-
ing to uniform conditions throughout the bed; (4) provide
ease of control; (5) can be designed for hazardous solids
and a wide range of temperatures (up to 1200 C), pressures
(up to 100 psig), residence times, and atmospheres; (6) can
operate on electricity, natural gas, fuel oil, thermal fluids,
steam, hot air, or hot water; (7) can process very fine and/or
low-density particles; and (8) provide very efficient emis-
sions control.
Under what conditions will the solid particles be fluid-
ized? At low gas velocities, solids are not fluidized but form
a fixed bed through which the gas flows upward with a
decrease in pressure due to friction and drag of the particles.
As the gas velocity is increased, the gas pressure drop across
the bed increases until the minimum fluidization velocity is
reached, where the pressure drop is equal to the weight of the
solids per unit cross-sectional area of the bed normal to gas
flow. At this point, the pressure drop is sufficient to support
the weight of the bed. The particle-levitation hydrodynamics
are similar to those of an airplane, which remains suspended
because the pressure below the wings is higher than that
above the wings. Further increases in gas velocity cause the

Stack
Fluidizing
chamber
Wet
feed
Feeder
Distributor
plate
Heat
source
Dry
Plenum
product
Fluidizing discharge
Air blower
inlet
(a) Single bed
Figure 18.10 Fluidized-bed dryers.
[From W.L. McCabe, J.C. Smith, and P. Harriott, Unit Operations of Chemical Engineering, 5th ed., McGraw-Hill, New York (1993) with permission.]
bed to expand with little or no increase in gas pressure drop.
Typically, fluidized-bed dryers are designed for gas velocities
of no more than twice the minimum required for fluidization.
That value depends on particle size and density, and gas den-
sity and viscosity. Superficial-gas velocities in fluidized-bed
dryers are from 0.5 to 5.0 ft/s, which provide stable, bubbling
fluidization. Higher velocities can lead to undesirable slug-
ging of large gas bubbles through the bed.
The capital and operating cost of a blower to provide suffi-
cient gas pressure for the pressure drops across the distributor
plate and the bed is substantial. Therefore, required solids-
residence time for drying is achieved by a shallow bed height
and a large chamber cross-sectional area. Fluidized-bed
heights can range from 0.5 to 5.0 ft or more, with chamber
diameters from 3 to 10 ft. However, chamber heights are
much greater than fluidized-bed heights, because it is desir-
able to provide at least 6 ft of free-board height above the top
surface of the fluidized bed, unless demisters are installed, so
that the larger dust particles can settle back into the bed
rather than be carried by the gas into the cyclone. Because of
intense mixing, temperatures of the gas and solids in a fluid-
ized bed are equal and uniform at the temperature of the dis-
charged gas and solids.
§18.1 Drying Equipment 735
Wet material
Clean gas
discharge
To cyclone
Dust
collector
Gas
Dry material
(b) Multiple beds
There is a substantial residence-time distribution for the
particles in the bed, which can be mitigated by baffles, multi-
staging, and mechanical agitators. Otherwise, a fraction of the
particles short-circuit from the feed inlet to the discharge duct
with little residence time and opportunity to dry. Another frac-
tion of the particles spend much more than the necessary resi-
dence time for complete drying. Thus, the nonuniform
moisture content of the product solids may not meet specifica-
tions. When the final moisture content is critical, it may be
advisable to smooth out the residence-time distribution by
using a more elaborate, multistage fluidized-bed dryer such as
the one shown in Figure 18.10b. Alternatively, the stages can
be arranged side by side horizontally. Starch dryers have been
fabricated with 20 such stages. Materials that are successfully
dried in fluidized-bed dryers include coal, sand, limestone,
iron ore, clay granules, granular fertilizer, granular desiccant,
sodium perborate, polyvinylchloride (PVC), starch, sugar,
coffee, sunflower seeds, and salt. Large fluidized-bed dryers
for coal and iron ore produce more than 500,000 lb/h of dried
material. For metallurgical applications and catalyst regenera-
tion, fluidized beds are frequently heated electrically and carry
price tags of from three to six million dollars depending on the
temperature and metallurgy requirements.
736 Chapter 18 Drying of Solids
Gas out
Solids out
Draft tube
Downcomer
Alternative
solids feed
Gas-distributor plate
Solid flow
Gas and solids feed Gas flow
Figure 18.11 Spouted-bed dryer.
Spouted-Bed Dryers
When wet, free-flowing particles are larger than 1 mm in
diameter but uniform in size and of low density, as in the
case of various grains, a spouted-bed dryer, shown in Figure
18.11, is a good choice, particularly when the required drying
time is more than just a few minutes. A high-velocity, hot gas
enters the bottom of the drying chamber, entrains particles,
and flows upward through a draft tube, above which the
cross-sectional area for gas flow is significantly increased in
a conical section, causing the gas velocity to decrease such that
the particles are released to an annular-downcomer region. A
fraction of the circulated solids are discharged from a duct in
the conical section. The gas exits at the top of the vessel from a
free-board region above the bed. For the drying of grains,
entrainment of dust in the exiting gas is minimal.
Pneumatic-Conveyor (Flash) Dryers
When only surface moisture must be evaporated from materi-
als that can be reduced to particles by a pulverizer, disinte-
gration mill, or other deagglomeration device, a pneumatic-
conveyor (gas-lift or flash) dryer is particularly desirable
when the material is temperature-sensitive, oxidizable, explo-
sive, and/or flammable. A flash dryer configuration is shown
in Figure 18.12. Wet solids are fed into a paddle-conveyor
mixer and dropped into a hammer mill, where solids are dis-
integrated. Air is pulled, by an exhaust fan, through a furnace
into the hammer mill, where the disintegrated solids are
Vent fan
Cyclone
Feeder
Flow
divider
Wet
feed Dry
product
Furnace Mixer
Hammer
mill
Figure 18.12 Pneumatic-conveyor (flash) dryer.
[From W.L. McCabe, J.C. Smith, and P. Harriott, Unit Operations of Chemi-
cal Engineering, 5th ed., McGraw-Hill, New York (1993) with permission.]
picked up and further deagglomerated into discrete particles
while being pneumatically conveyed upward at high velocity
in a duct where much of the drying takes place. The particle-
gas mixture is separated in a cyclone separator, from which
the solids are discharged. Because the particles travel upward
in the drying duct at a velocity almost equal to that of the gas,
a residence time of less than 5 s is provided. If additional time
is needed, up to 30 s can be achieved by partial recycle of the
solids leaving the cyclone separator. Recycle of solids is also
useful for the disintegration of materials that are sticky or
pasty. The deagglomerated particle sizes are from 30 to
300 mesh. If the particles are crystalline or friable, they may
be subject to excessive breakage. Pneumatic conveying veloc-
ities range from 10 to 30 m/s, usually about 3 m/s greater than
the terminal (free-fall) velocity of the largest particle to be
conveyed out of the disintegrator. The distribution of remain-
ing moisture in the product particles can be wide because of
the distribution of particle-residence times. However, surface
drying is rapid because inlet gas temperatures as high as
1,500 F can be employed. Nevertheless, because: (1) the par-
ticles flow cocurrently with the gas, (2) the gas temperature
decreases significantly, and (3) the particle-residence time is
short and evaporation of moisture is incomplete, the particles
do not attain temperatures greater than about 200 F.

Large flash dryers are provided with pneumatic-conveying
dryer ducts 1 m in diameter and 12 m high, with water-
evaporation capacities up to 36,000 lb/h. Compared to many
other dryers, they have small floor-area requirements and are
used for drying filter cakes, centrifuge cakes and slurries,
yeast cakes, whey, starch, sewage sludge, gypsum, fruit pulp,
copper sulfate, clay, coal, chicken droppings, adipic acid,
polystyrene beads, ammonium sulfate, and hexamethylene
tetramine.
Spray Dryers
When solutions, slurries, or pumpable pastes—containing
more than 50 wt% moisture, at rates greater than 1,000 lb/h—
are to be dried, a spray dryer should be considered. In the con-
figuration in Figure 18.13a the drying chamber has a conical-
shaped bottom section with a top diameter that may be nearly
equal to the chamber height. Feed is pumped to the top center
of the chamber, where it is dispersed into droplets or particles
from 2 to 2,000 mm by any of three types of atomizers:
(1) single-fluid pressure nozzles, (2) pneumatic nozzles, and
(3) centrifugal disks or spray wheels. Hot gas enters the cham-
ber, causing moisture in the atomized feed to rapidly evapo-
rate. Gas flows cocurrently to the solids, and dried solids and
gas are either partially separated in the chamber, followed by
removal of dust from the gas by a cyclone separator, or, as
shown in Figure 18.13a, are sent together to a cyclone separa-
tor, bag filter, or other gas–solid separator. The hot gas can be
moved by a fan.
In many respects, spray dryers are similar in operating
conditions to a pneumatic-conveyor dryer because particles
Fan
Spray dryer
Furnace Bag collector
Feed pump
(a) Process system
(b) Centrifugal disk atomizer
Figure 18.13 Spray dryer.
§18.1 Drying Equipment 737
are small, entering gas temperature can be high, residence
time of the particles is short, mainly surface moisture is
removed, and temperature-sensitive materials can be
handled. However, a unique feature of spray dryers is their
ability, with some materials such as dyes, foods, and deter-
gents, to produce, from a solution, rounded porous particles
of fairly uniform size that can be rapidly dissolved or reacted
in subsequent applications.
Although residence times are less than 5 s if only surface
moisture is removed, residence times of up to 30 s can be
provided for evaporating internal moisture. Spray drying is
also unique in that it combines, into one compact piece
of equipment, evaporation, crystallization or precipitation,
filtration or centrifugation, size reduction, classification,
and drying.
A critical part of a spray dryer is the atomizer. Each of the
three atomizer types has advantages and disadvantages.
Pneumatic (two-fluid) nozzles impinge the gas on the feed at
relatively low pressures of up to 100 psig, but are not efficient
at high capacities. Consequently, they find applications only
in pilot plants and low-capacity commercial plants. Excep-
tions are the dispersion of stringy and fibrous materials,
thick pastes, certain filter cakes, and polymer solutions be-
cause with high atomizing gas-to-feed ratios, small particles
are produced.
Pressure (single-fluid) nozzles, with orifice diameters of
0.012–0.15 inch, require solution inlet pressures of 300–
4,000 psig to achieve breakup of the feed stream. These noz-
zles can deliver the narrowest range of droplet sizes, but the
droplets are the largest delivered by the three types of atom-
izers, and multiple nozzles are required in large-diameter
spray dryers. Also, orifice wear and plugging can be prob-
lems with some feeds. Because the spray is largely down-
ward, chambers are relatively slender and tall, with height-
to-diameter ratios of 4–5.
The centrifugal disks (spray wheels) shown in Figure
18.13b handle solutions or slurries, delivering thin sheets of
feed that break up into small droplets in a nearly radial direc-
tion at high capacities. Disks have the largest-diameter spray
pattern and therefore require the largest-diameter drying
chambers to prevent particles from striking the chamber wall
while in a sticky state.
Centrifugal disks range in diameter from a few inches up
to 32 inches in large units. Disks spin at 3,000–50,000 rpm,
and can operate over a range of feed and rotation rates with-
out significantly affecting the particle-size distribution that
occurs with variation of the feed pressure or by enlargement
of or other damage to the orifice of a pressure nozzle.
Industrial spray-dryer diameters are large, with 8–30 ft
being common. Evaporation rates of up to 2,600 lb/h have
been achieved in an 18-ft-diameter by 18-ft-high spray dryer
equipped with a centrifugal-disk atomizer and supplied with
an aqueous feed solution of 7 wt% dissolved solids and
11,000 cfm of air at 600 F. Larger spray dryers can evaporate
up to 15,000 lb/h. Solutions, slurries, and pastes of the follow-
ing materials are spray-dried: detergents, blood, milk, eggs,
starch, yeast, zinc sulfate, lignin, aluminum hydroxide, silica
gel, magnesium chloride, manganese sulfate, aluminum
738 Chapter 18 Drying of Solids

sulfate, urea resin, sodium sulfide, coffee extract, tanning Drum dryers are used to dry a wide variety of materials,
extract, color pigments, tea, tomato juice, polymer resins, and including brewer’s yeast, potatoes, skim milk, malted milk,
ceramics. coffee, tanning extract, and vegetable glue; slurries of
Mg(OH)2, Fe(OH)2, and CaCO3; and solutions of sodium
acetate, Na2SO4, Na 2HPO4, CrSO4, and various organic com-
Drum Dryers
pounds. Drum dryers belong to a class of hot-cylinder dryers.
Approximately 100 years ago and well before the 1920s, Units with large numbers of cylinders in series and parallel
when spray dryers were introduced for drying milk and deter- are used to dry continuous sheets of woven fabrics and paper
gent solutions, drum dryers were in use to process solutions, pulp at evaporation rates of about 10 kg/h-m2.
slurries, and pastes with indirect heat. The first such dryer
was the double-drum dryer, shown in Figure 18.14a, which is
§18.1.4 Other Dryers
still the most versatile and widely used drum dryer. It consists
of two metal, cylindrical drums of identical size (1–5 ft diam- A number of other dryers have been developed for special
eter by 1.5–12 ft long), mounted side by side. One drum is situations. These use infrared radiant energy, generation of
movable horizontally so that the distance between the two heat within the solid by dielectric drying using radio or
drums (the nip) can be adjusted. The drums are heated on the microwave frequencies, and freeze-drying by sublimation of
inside by condensing steam at pressures as high as 12 atm. frozen moisture.
The feed enters at the top from a perforated pipe that runs the
length of the two drums or from a pipe that swings like a pen-
dulum from end-to-end of the drums. The drums are rotated Infrared Drying
toward each other at the top, as shown, causing the feed to
In direct-heat dryers, the transfer of heat by convection from
form, on the hot surface of the drums, a clinging coating,
hot gases to the wet material is often inadvertently supple-
whose thickness is controlled by the nip. As the drums rotate,
mented by thermal radiation from hot surfaces that surround
heat is transferred to the coating, causing it to dry. If the
the material. This radiant-heat contribution is usually minor,
moisture is water, it exits as steam through a vapor hood. If
and ignored. For the drying of certain films, sheets, and coat-
the moisture is a solvent or if the solid is dustable, a dryer
ings, however, use of thermal radiation as the major source of
enclosure can be provided. When the coating has made about
heat is a proven technology.
3/4 of a complete rotation, it is scraped off the drum surfaces
Radiant energy is released from matter as a result of oscil-
by doctor blades that run the length of the drums. Dried mate-
lations and transitions of its electrons. For gases and transpar-
rial falls off—as surface powder, chips, or, more commonly,
ent solids and liquids, radiation can be emitted from
as flakes, which are 1–3 mm thick—into conveyors. By
throughout the volume of the matter. For opaque solids and
adjusting (1) drum-rotation rate from 1 to 30 rpm; (2) drum-
liquids, this internal radiation is quickly absorbed by adjoin-
surface temperature, usually just a few degrees below the
ing molecules so that the net transfer of energy by radiation is
inside, condensing-steam temperature; (3) feed temperature;
only from the surface. Of great importance in radiation heat
and (4) coating thickness, the moisture content of the dried
transfer for drying is the transfer of radiation from a hot, opa-
material can be controlled. Drying times are 3–20 s. Perform-
que surface through a nonabsorbing gas or vacuum to the
ance data given by Walas [3] show capacities of double-drum
material being dried. This transfer can be viewed as the prop-
dryers to be in the range of 1–60 kg of dried product/h-m2 of
agation of discrete photons (quanta) and/or as the propaga-
drum surface for feed moisture contents of 10–90 wt%.
tion of electromagnetic waves, consisting, as shown in
For drum dryers to be effective, the coating must adhere to
Figure 18.15a, of electric and magnetic fields that oscillate at
the drum surface, which is often chrome-plated. When neces-
right angles to each other and to the direction in which the
sary, other drum and feeding arrangements can be employed.
radiation travels. As shown in the electromagnetic spectrum
The twin-drum dryer with top feed shown in Figure 18.14b is
of Figure 18.15b, the wavelength, l, of the radiation, which
not influenced by drum spacing because the drums rotate
depends on the manner in which it is generated, covers an
away from each other at the top. Thicker coatings can then
exceedingly wide range, from gamma rays of 10 8 mm to
be formed, and materials like inorganic crystals that might
long radio waves of 1010 mm. Regardless of the wavelength,
score or cause damage to closely spaced drums can be
all radiation waves travel at the speed of light, c, which for a
processed. To improve the likelihood of the feed adhering
vacuum is 2.998 108 m/s. Accordingly, a relationship exists
to the drums, feed may be splashed onto the surface of the
between frequency of the wave, y, and its wavelength, l:
drum, as in Figure 18.14c.
For very viscous solutions or pastes that might cause undue y ¼ c=l ð18-1Þ
pressure to be put on the surfaces of a double-drum dryer, a
single-drum dryer can be used. The coating can be applied by The frequency is usually expressed in Hz, which is one cycle/s.
using a top feed with applicator rolls, as shown in Figure The energy transmitted by the wave depends on its frequency
18.14d. If a porous product (e.g., malted milk) is desired or if and is expressed in terms of the energy, E, of a photon by
the material is temperature-sensitive, a single-drum or double- E ¼ hn ð18-2Þ
drum dryer, as shown in Figure 18.14e, can be enclosed so that
34
a vacuum can be pulled to reduce the boiling point. where h ¼ Planck’s constant ¼ 6.62608 10 J-s.
 §18.1 Drying Equipment 739

Feed pipe

Knife Knife Drum Drum

Knife Knife

Steam-heated Steam-heated
drum drum

(c) Twin-drum dryer with splash feed

Conveyor Conveyor
(a) Double-drum dryer
Feed
pipe
Applicator
roll

Drum
Drum Drum Knife

Knife Knife

(b) Twin-drum dryer with top feed

(d) Single-drum dryer with applicator feed

Vapor outlet

Pendulum
feed

Knife Knife

Drum Drum

Conveyor Conveyor

Manhole

(e) Vacuum double-drum dryer

Figure 18.14 Drum dryers.
740 Chapter 18 Drying of Solids

y E H Sources of infrared radiant heat at surface temperatures in

Eo
the range of 600–2,500 K are electrically heated metal-sheath
x rods, quartz tubes, and quartz lamps; and ceramic-enclosed
gas burners. When the radiant energy reaches the material
Ho
being dried, it is absorbed at the surface, from which it
z E is electric component H is magnetic component is transferred into the interior by conduction. In this respect,
Eo and Ho are amplitudes infrared radiant heat transfer to the surface is much like con-
(a) Electromagnetic wave vective heat transfer. However, if the effective thermal con-
ductivity of the material is low, the surface temperature may
Wave length Type of radiation rise to an undesirable value, particularly if high-temperature,
µm Å infrared-radiation sources are used. Applications of growing
10–8 10–4 interest include drying of paper, paints, enamels, inks, glue-
10–7 10–3 on flaps, and textiles. Continuous infrared dryers are more
10–6 10–2 Gamma rays common than batch infrared dryers.
10–5 10–1
10–4 1 X-rays
10–3 10 Dielectric Drying
10–2 102 Ultraviolet
10–1 103
In contrast to infrared drying, dielectric drying involves the
Visible low-frequency, long-wavelength end of the electromagnetic
4
1 10
Thermal spectrum of Figure 18.15b, where radio waves and micro-
10 105 radiation
Infrared waves reside. With nonelectrically conducting materials,
102 106
heat is not absorbed at the surface but is generated through-
103 107
out the material, reducing the importance of heat conduction
104 108
within the material and, thus, making this type of drying
105 109 unique and making it possible to control the rate of energy
6 Radar,
10 1010 dissipation in the material over a wide range. Other advan-
television,
107 1011 FM broadcast tages over more conventional drying methods include: (1)
8
10 1012 efficiency of energy usage because the energy dissipation
9
10 1013 occurs mainly in the moisture rather than in the solid mate-
10
10 1014 rial; (2) operation at low temperatures, thus avoiding high
material surface temperatures; and (3) more rapid drying.
10–10m = 1 angstrom Dielectric drying is particularly useful for preheating materi-
10–6m = 1 micron
als and for removing the final traces of internal moisture.
(b) Electromagnetic spectrum Radio frequency (RF) drying is confined to frequencies, y,
Figure 18.15 Radiation. between 1 and 150 MHz (l ¼ 3 108 to 2 106 mm or 300
to 2 m), while microwave drying utilizes frequencies from
300 MHz to 300 GHz (l ¼ 106 to 103 mm or 1 m to 1 mm).
A solid, opaque surface can emit infrared radiation by vir- By international agreement, microwave drying is done at
tue of its temperature. This type of radiation is invisible and only 915 and 2,450 MHz, as discussed by Mujumdar [1]. For
has a wavelength, as shown in Figure 18.15b, in the range of RF drying, the U.S. Federal Communications Commission
0.75–300 mm. If the surface emitting the radiation is a so- (FCC) has reserved frequencies of 13.56 MHz 0.05%,
called blackbody, b, such that the maximum amount of radia- 27.12 MHz .60%, and 40.68 MHz 0.05%. Equipment
tion will be emitted, that amount will be distributed over a for dielectric drying consists of an energy generator and an
range of wavelengths, depending on the temperature, as gov- applicator. A generator is used to boost 50–60 Hz line voltage
erned by the Planck distribution, which, in terms of radiant to the much higher values quoted above. A negative-grid tri-
heat leaving diffusely from a unit area of surface, is ode tube is used with dielectric systems, while magnetron or
klystron tubes are used with microwave systems. Dielectric
El;b ¼ C 1 =fl5 ½expðC 2 =lTÞ 1g ð18-3Þ
energy is usually applied by electrodes of various types and
2
where the units of El,b are W/m -mm, C1 ¼ 3:742 108 , shapes, between which is placed the material to be dried.
C2 ¼ 1:439 104 , T is in K, and l is in mm. When (18-3) is Microwave systems often use hollow, rectangular, metallic
integrated over the entire range of wavelengths, the result is waveguides.
the Stefan–Boltzmann equation, RF systems are used to dry bulky materials such as lum-
ber, ceramic monoliths, foam rubber, breakfast cereals, dog
Eb ¼ sT 4 ð18-4Þ
biscuits, crackers, biscuits, and cookies, as well as films,
2 4
where s ¼ 5:67051 10 8 W/m -K and T is in K. Thus, as coatings, and materials such as paper, inks, adhesives, tex-
the temperature of the infrared heat source is increased, the tiles, and penicillin, where high surface temperatures must be
rate of heat transfer increases exponentially. avoided. Ceramic catalytic-converter extrusions are dried

Gate valve
Vapor constriction plate
Forward
vacuum
locks
1 2 3 4
Food trolley moving Vacuum joint
between fixed heating (tunnel sections)
platens
quickly and uniformly by RF drying. Microwave systems are
used to dry pasta, onions, seaweed, baseball bats, potato chips,
pharmaceuticals, ceramic filters, and sand casting molds.
Freeze-Drying
In freeze-drying (lyophilization), moisture in the feed is first
frozen, by cooling, and then sublimed by conductive, convec-
tive, and/or radiant heating. Because the structure and prop-
erties of solid material to be dried are hardly altered by
freeze-drying, it has been adapted widely to the drying of
biological materials, pharmaceuticals, and foodstuffs. Prod-
ucts of freeze-drying are porous and nonshrunken. When
foodstuffs are dehydrated by freeze-drying and then stored
under a dry, inert gas, they evade deterioration almost indefi-
nitely and can be rehydrated almost perfectly to their original
state for later consumption. The first major application of
freeze-drying was for the preservation of blood plasma
during World War II.
When the moisture is water, the material must be cooled
to at least 0 C to freeze the water if it is free, and even lower
if the water is dissolved in the material. Most freeze-drying is
conducted at 10 C or lower. At this temperature, ice has a
vapor pressure of only 2 torr; therefore, freeze-drying must
be conducted under a high vacuum. Heat for sublimation is
transferred from the heat source to the material under con-
trolled conditions so that the moisture does not reach the
melting point. In some cases, an even lower temperature,
called the scorch point, must not be exceeded, or degradation
of the material will occur. During the drying period, which
may take 20 hours, resistance to heat transfer increases
because an interface develops between the porous freeze-
dried layer and the frozen material, which gradually recedes
into the material.
For small quantities of biological and pharmaceutical
materials, freeze-drying is conducted batchwise on trays in
vacuum cabinets, where the drying step follows the freezing.
The sublimed ice is desublimed on a cold metal plate that
§18.2 Psychrometry 741
Electrical
controller
Vacuum
gauge
Aft
vacuum
locks
5
Condensing
chamber
To vacuum pump
Figure 18.16 Tunnel freeze-dryer.
resides either inside the cabinet and adjacent to the trays or
in a separate, adjoining vessel. During sublimation, heat
transfer is usually by conduction from the bottom- and side-
tray surfaces, which contain coils or passages through which a
heating fluid, e.g., vacuum steam, passes. For large quantities
of foodstuffs, continuous freeze-drying can be employed, as
shown in Figure 18.16, using trays of prefrozen materials
transported through a tunnel past fixed heating platens, with
vacuum locks at either end. With granular materials, drying
times of less than 1 h can be achieved. Continuous freeze-
drying of small particles can also be accomplished rapidly in
a fluidized bed, where heat transfer for sublimation is by con-
vection and radiation. However, bed turbulence may result in
particle breakage and dusting. In some cases, freeze-drying
can utilize infrared and microwave heating. Freeze-drying is
used for the sublimation of moisture from seafood, meat,
vegetables, fruits, coffee, concentrated beverages, pharma-
ceuticals, blood plasma, and biological materials.
§18.2 PSYCHROMETRY
If moisture is to evaporate from a wet solid, it must be heated
to a temperature at which its vapor pressure exceeds the par-
tial pressure of the moisture in the gas in contact with the wet
solid. In an indirect-heat dryer, where little or no gas is used
to carry away the moisture as vapor, the partial pressure of
the moisture approaches the total pressure, and the tempera-
ture of evaporation approaches the boiling point of the mois-
ture at the prevailing pressure, as long as the moisture is free
liquid at the surface of the solid. If the moisture interface
recedes into the solid, a temperature above the boiling point
is necessary at the solid–gas interface to transfer the heat for
evaporation to the liquid–gas interface. If the moisture level
drops to a point where it is entirely sorbed, its vapor pressure
is less than the pure vapor pressure and an even higher tem-
perature is required to evaporate it. In a direct-heat dryer,
similar situations occur, except that the temperature at which
moisture evaporates depends on the gas-moisture content.
742 Chapter 18 Drying of Solids
§18.2.1 Psychrometric Chart
Calculations involving the properties of moisture–gas mix-
tures for application to drying are most conveniently carried
out with psychrometric charts. A typical chart, given in
Figure 18.17, is that for air–water vapor mixtures at 1-atm
total pressure. Included in this chart are properties that are
listed and defined in Table 18.4, which applies to general
moisture–gas mixtures that obey the ideal-gas law. The defi-
nitions given by (18-5)–(18-9) for humidity and by (18-10)–
(18-12) for humid volume, humid heat, and enthalpy are in
terms of a unit mass or mole of moisture-free gas. For a water
vapor–air mixture, the term ‘‘dry air’’ is substituted for
Figure 18.17 Psychrometric
(humidity) chart for air–water
at 1 atm.
‘‘moisture-free gas.’’ The following example, which makes
use of Figure 18.17 and Table 18.4, illustrates the use of the
dry-air or ‘‘dry’’ basis.
EXAMPLE 18.1 Use of the Psychrometric Chart.
Air at 131 F and 1 atm enters a direct-heat dryer with a humidity, ,
of 0.03 lb H2O (A)/lb H2O-free air (B). Determine by the psychro-
metric chart of Figure 18.17 and the relationships of Table 18.4:
(a) molal humidity, m ; (b) saturation humidity, s ; (c) relative
humidity, R ; (d) percentage humidity, P ; (e) humid volume, yH;
(f) humid (specific) heat, Cs; and (g) enthalpy, H.
 §18.2 Psychrometry 743

Table 18.4 Definitions of Quantities Useful in Psychrometry: A = moisture; B = moisture-free gas, ideal-gas conditions

Quantity Definition Relationship

Absolute, mass humidity Moisture content of a gas by mass M A pA (18-5)

¼
M B ðP pA Þ

pA
Molal humidity Moisture content of a gas by mols m ¼ (18-6)
P pA

M A PsA
Saturation humidity Humidity at saturation s ¼ (18-7)
M B P PsA

pA
Relative humidity (relative saturation as Ratio of partial pressure of moisture to partial R ¼ 100% (18-8)
PsA
a percent) pressure of moisture at saturation

Percentage humidity (percent Ratio of humidity to humidity at saturation P ¼ 100% (18-9)

s
saturation)

RT 1
Humid volume Volume of moisture–gas mixture per unit mass of yH ¼ þ (18-10)
P M B MA
moisture-free gas
Humid heat Specific heat of moisture–gas mixture per unit mass C s ¼ ðCP ÞB þ ðC P ÞA (18-11)
of moisture-free gas
Total enthalpy Enthalpy of moisture–gas mixture per unit mass of H ¼ C s ðT T o Þ þ DH vap
o (18-12)
moisture-free gas referred to temperature, To
Dew-point temperature Temperature at which moisture begins to condense Tdew
when mixture is cooled
Dry-bulb temperature Temperature of mixture Td
Wet-bulb temperature Steady-state temperature attained by a wet-bulb Tw
thermometer
Adiabatic-saturation temperature Temperature attained when a gas is saturated with Ts
moisture in an adiabatic process

Solution
atm-ft3
(e) From (18-10), for R ¼ 0:730 ; T ¼ 131 460 ¼ 591 R,
At 131 F, the vapor pressure of water is 118 torr ¼ 0.155 atm. lbmol- R
0:730ð591Þ 1 0:03
(a) Combining (18-5) and (18-6), yH ¼ þ ¼ 15:6 ft3 =lb dry air
1 28:97 18:02
MB M air 28:97 lb mol H2 O
m ¼ ¼ ¼ ð0:03Þ ¼ 0:048 which agrees with Figure 18.17.
MA M H2 O 18:02 lb mol dry air
(f) From (18-11), using ðC P Þair ¼ 0:24 Btu/lb- F and ðC P Þsteam ¼
(b) From (18-7), 0:45 Btu/lb- F,
18:02 0:155 lb H2 O Btu
s ¼ ¼ 0:114 C s ¼ 0:24 þ ð0:45Þð0:03Þ ¼ 0:254
28:97 1 0:155 lb dry air lb dry air
(g) Equation (18-12) assumes that the enthalpy datum refers to air
(c) From a rearrangement of (18-6), as a gas and water as a liquid. Taking T o ¼ 32 F and DH vap
o ¼
1075 Btu=lb, (18-12) gives
P m ð1Þð0:048Þ
pH2 O ¼ ¼ ¼ 0:0458 atm H ¼ 0:254ð131 32Þ þ 1; 075ð0:03Þ ¼ 57:4 Btu/lb dry air.
1þ m 1 þ 0:048

0:0458
From (18-8), R ¼ 100 ¼ 29:5%
0:155

The same result is obtained from Figure 18.17. EXAMPLE 18.2 Humidity for Benzene as the Moisture.
In a dryer where benzene (A) is evaporated from a solid, nitrogen
0:03
(d) From (18-9), P ¼ 100 ¼ 26:3% gas (B) at 50 F and 1.2 atm has a relative humidity for benzene of
0:114 35%. Determine: (a) benzene partial pressure if its vapor pressure at
744 Chapter 18 Drying of Solids

50 F ¼ 45.6 torr, (b) humidity of the nitrogen–benzene mixture, (c) the solid, this temperature of evaporation is called the wet-
saturation humidity of the mixture, and (d) percentage humidity of bulb temperature, Tw, because it can be measured by cover-
the mixture. ing a thermometer bulb with a wick saturated with the liquid
being evaporated and passing a partially saturated gas past
Solution the wick, as indicated in Figure 18.18a.
R ¼ 35%; P ¼ 1:2 atm ¼ 912 torr; In Figure 18.18b, where the wetted wick is replaced by an
M A ¼ 78:1; M B ¼ 28; Psbenzene ¼ 45:6 torr incremental amount of wet solid, assume the heat-transfer
area ¼ mass-transfer area ¼ A. At steady state, the rate of
(a) From (18-8), convective heat transfer from the gas to the wet solid is given
Psbenzene R ð45:6Þð35Þ by Newton’s law of cooling:
pbenzene ¼ ¼ ¼ 16 torr
100 100 Q ¼ hðT T w ÞA ð18-13Þ
(b) From (18-5), The molar rate of mass transfer of evaporated moisture from
78:1 16 lb benzene the wet surface of the solid, A, is
¼ ¼ 0:050
28 912 16 lb dry nitrogen
ky yAW yA A
nA ¼ ð18-14Þ
(c) From (18-7), ð1 yA ÞLM
78:1 45:6 lb benzene An enthalpy balance on the moisture evaporated and heated
s ¼ ¼ 0:147
28 912 45:6 lb dry nitrogen to the gas temperature couples the heat- and mass-transfer
0:050 equations to give
(d) From (99), P ¼ 100 ¼ 34%
0:147
Q ¼ nA M A DH vap
w þ ðC P ÞA ðT T wÞ ð18-15Þ

To obtain a simplified relationship for the coupling in terms of

T and , assume that the mole fraction of moisture in the bulk
gas and at the wet solid–gas interface is small. Then, the bulk-
§18.2.2 Wet-Bulb Temperature
flow effect in (18-14) becomes ð1 yA ÞLM 1:0. Also, from
The temperature at which moisture evaporates in a direct- (18-5), replacing pA with yAP,
heat dryer is difficult to determine and varies from the dryer
MA MB
inlet to the dryer outlet. When the dryer operates isobarically yA ¼ ð18-16Þ
1 MA
and adiabatically, with all energy for moisture evaporation þ
supplied from the hot gas by convective heat transfer, with MB MA
no energy required for heating the wet solid to the evapora- If the latent heat in (18-15) is much greater than the sensi-
tion temperature, use of simplified heat- and mass-transfer ble heat,
equations leads to an expression for the temperature of evap-
oration at a particular location in a dryer operating under DH vap
w þ ðC P Þ A ðT T wÞ DH vap
w ð18-17Þ
continuous, steady-state conditions, or at a particular time in
Simplifying, and combining (18-13) to (18-16),
a batch dryer cycle.
If it is further assumed that the moisture being evaporated ky M B DH vap
w
Tw ¼ T ð w Þ ð18-18Þ
is free liquid exerting its full vapor pressure at the surface of h

TW TW

Make-up liquid

Gas Gas
Figure 18.18 Wet-bulb
temperature.
Temperature, T Temperature, T
[From W.L. McCabe, J.C. Smith,
Humidity, Humidity, and P. Harriott, Unit Operations
of Chemical Engineering, 5th ed.,
McGraw-Hill, New York (1993)
(a) Wick (b) Wet solid with permission.]
 §18.2 Psychrometry 745

Table 18.5 Lewis Number for Liquids Evaporating into Air at Saturated gas
25 C Ts, s, E + mB

(NLe)2/3 ¼
Liquid
Liquid NLe Psychrometric Ratio
TL, mL
Isobaric,
Benzene 2.44 1.812 adiabatic
Carbon tetrachloride 2.67 1.923 saturator
Chloroform 3.08 2.114 Partially saturated
Ethyl acetate 2.58 1.880 gas
Ethylene tetrachloride 3.05 2.101 T, , mB

Metaxylene 3.18 2.165 Excess liquid

Methanol 1.37 1.233 Ts, mL – E
Propanol 1.85 1.506
Figure 18.19 Adiabatic saturation of a gas with a liquid.
Toluene 2.64 1.908
Water 0.855 0.901

temperature Ts. An enthalpy balance using (18-1) for

If the Chilton–Colburn analogy for mass and heat transfer enthalpy of the gas phase, but with the reference temperature
applies, then from (3-165) and (3-228), To ¼ Ts (so as to simplify the balance), gives
ky M 2=3 h 2=3 mL ðC PA ÞL ðT L T s Þ þ mB C sin ðT T s Þ þ DH vap
N ¼ N ð18-19Þ s
G Sc C P G Pr
¼ mB C sout ðT s T s Þ þ DH vap
s s ð18-23Þ
If M MB, (18-19) becomes
2=3 þðmL E Þ þ ðCPA ÞL ðT s T sÞ
ky M B 1 1
¼ ð18-20Þ Assume that the sensible heat required to heat the liquid from
h ðCP ÞB N Le
TL to Ts is negligible. Then (18-23) simplifies to an equation
where N Le ¼ Lewis number ¼ N Sc =N Pr ð18-21Þ for the adiabatic-saturation temperature:
The reciprocal of the Lewis number is referred to as the
Luikov number, NLu. Substituting (18-20) into (18-18), DH vap
s
Ts ¼ T ð s Þ ð18-24Þ
2=3
ðC s Þin
DH vap
w 1
Tw ¼ T ð w Þ ð18-22Þ
CP N Le Equation (18-24) can be used to determine gas tempera-
In Equation (18-18), which defines the wet-bulb tempera- tures and humidities between T and Ts, and and s , if sensi-
ture, or in (18-22), it is important to note that w is the satu- ble heat for the liquid is ignored. If (18-24) is compared to
ration humidity at temperature Tw. (18-18), it is seen that wet-bulb and adiabatic-saturation tem-
By coincidence, (NLe)2/3 for air–water vapor at 25 C is peratures are equal if
close to 1.0 (actually, 0.901). For the air–organic-vapor sys- h
psychrometric ratio ¼ Csin ðN Le Þ2=3 ¼ 1
tems listed in Table 18.5, which is taken from Keey [7], (NLe)2/3 ky M B
is much less than 1.0. As shown in the next section, (NLe)2/3 has
an impact on the variation with location or time of the tempera- For the air–water system, as shown by Lewis [8] and referred
ture of moisture evaporation in a direct-heat dryer. to as the Lewis relation, this is almost the case, with
h
0:216 Btu=lb- F
ky M B
§18.2.3 Adiabatic-Saturation Temperature
compared to Cs 0.24 Btu/lb- F. Normally, a small amount
To determine the change in the wet-bulb temperature of a wet of thermal radiation heat transfer to the wet solid supple-
solid, as it dries in the dryer, it is necessary to consider ments the convective heat transfer, so if h in the psychromet-
changes in the temperature and humidity of the gas as it ric ratio is replaced by (hc þ hr), hr is an effective heat-
cools, because of heat transfer to the wet solid and transfer of transfer coefficient for thermal radiation, and the corrected
moisture from the wet solid. A simplified relationship be- ratio is almost identical to the humid heat at low humidities.
tween gas temperature and humidity can be derived by con- Accordingly, for the air–water system, the wet-bulb tempera-
sidering the adiabatic saturation of a gas with an excess of ture is set equal to the adiabatic-saturation temperature
liquid. Referring to Figure 18.19, partially saturated gas at and only one family of lines is shown on the psychrometric
temperature T, humidity , and mass flow rate mB (dry basis), chart. From (18-24), these lines have a negative slope of
together with liquid at temperature TL and mass flow rate mL, DH vap
s =ðC s Þm . The following example illustrates use of the
enters an isobaric and adiabatic chamber where a fraction of chart in Figure 18.17 to determine the relationship between
the liquid, (E/ml), is vaporized to saturate the gas. The gas gas (dry-bulb) temperature, wet-bulb temperature, and
and excess liquid leave the chamber in equilibrium at humidity. The example also illustrates use of (18-24).
746 Chapter 18 Drying of Solids

EXAMPLE 18.3 Wet-Bulb and Adiabatic-Saturation 18:02 45 lb H2 O

Temperatures. s ¼ ¼ 0:0391
28:97 760 45 lb dry air
For the conditions of Example 18.1, determine the wet-bulb temper- 1038:7
T s ¼ 131 ð0:0391 0:03Þ ¼ 94 F
ature, assuming it is equal to the adiabatic-saturation temperature, 0:254
by using (a) the psychrometric chart and (b) (18-24). As can be seen, this iterative calculation is very sensitive. By inter-
polation, Ts ¼ 95.5 C, which is close to the value of 96.5 F read
Solution from Figure 18.17.
(a) In Figure 18.17, the point T ¼ Td ¼ 131 F and ¼ 0:03 lb H2O/
lb dry air is plotted. This point lies just above the adiabatic-satu- For systems other than air–water, the Lewis relation does
ration-temperature line of 95 F at about 96.5 F. not hold, and it is common to employ the psychrometric
(b) Equation (18-24) involves an iterative calculation to determine ratio, h/(kyCsMB). Calculated values of this ratio, using the
Tw ¼ Ts because DH vap Chilton–Colburn analogy, are included in Table 18.5. Com-
s and s are unknown.
paring (18-18) for wet-bulb temperatures to (18-24) for adia-
Assume: Ts ¼ 95 F, DH vap s ¼ 1039:8 Btu/lb from steam tables, batic-saturation temperatures, it can be noted that since the
ðCs Þin ¼ 0:254 Btu/lb from Example 18.1, and PsA ¼ 42:2 torr ¼
inverse of the psychrometric ratio for all air–organic vapor
vapor pressure of water from Example 18.1.
systems in Table 18.5 is less than 1, these systems will have
18:02 42:2 lb H2 O slopes of adiabatic-saturation lines that are less than slopes of
From (18-7), s ¼ ¼ 0:0366
28:97 760 42:2 lb dry air wet-bulb lines. An example is the psychrometric chart for
air–toluene at 1 atm shown in Figure 18.20. Use of this chart
1039:8
From (18-24), T s ¼ 131 ð0:0366 0:03Þ ¼ 104 F is illustrated in the next example.
0:254
(So, try again.) EXAMPLE 18.4 Psychrometric Chart for Air-Toluene.
Assume:
T s ¼ 97 F; DH vap ¼ 1038:7 Btu=lb; and PsA ¼ 45 torr; then Air is used to dry a solid wet with toluene at 1 atm. At a location
s
where the air has a temperature of 140 F and a relative humidity of

Humid heat – B.t.u./(lb.dry air)(°F) Percent relative humidity

0.25 0.30 0.35 90 70 50 40 30 20 10
0.40
e
ity lin

23
22
humid

0.35 Lat
Specific volume – cu. ft./lb. of dry air

21 en
th
eat
ated

20 of
190 va po
Absolute humidity-lb. toluene vapor/lb. of dry air

riza
Satur

19 ti
Latent heat of evaporation, B.t.u./lb.

0.30 on
vs.
18 te ir
mp dry a
. e of
180 olum
17 ific v
0.25 Spec
16 Ad
iab
170 15 at 5
ic
e s atu
14 0.20 olum ra
t ed v tio
ura n
13 Sat lin
es
160
idity

12
Li
ne

0.15
hum

so

11
fc
on
abs.

150
st
nta
vs.

0.10
w
eat

et
-b
id h

ul
b
te
Hum

0.05
pe
ra
tu
re

0
30 40 60 80 100 120 140 160 180 200 220 240
Dry-bulb temperature, °F
Figure 18.20 Psychrometric (humidity) chart for air–toluene at 1 atm.
[From Perry’s Chemical Engineers’ Handbook, 6th ed., R.H. Perry, D.W. Green, and J.O. Maloney, Eds., McGraw-Hill, New York (1984) with permission.]
 §18.2 Psychrometry 747

10%, determine the humidity, the adiabatic-saturation temperature, 150

and the wet-bulb temperature. 140

130
Solution
120 Gas

Temperature, °F
From Figure 18.20, the humidity is 0.062 lb toluene/lb dry air at the
110
intersection point of a vertical temperature line and a curved percent
relative-humidity line. By following an interpolated adiabatic- 100
saturation line from that intersection point to the saturated-humidity Wet solid (at wet-bulb temperature)
90
line, Ts ¼ 83 F. By following an interpolated wet-bulb line from the
intersection point to the saturated-humidity line, Tw ¼ 92 F. Thus, 80
the wet-bulb temperature is higher than the adiabatic-saturation tem- 70
perature. In the next section, this causes a reduction in the driving
60
force available for heat transfer during the drying of solids wet with
organic moisture. 50
0 1 2 3 4 5 6 7 8 9 10
Distance through dryer, ft
(a) Water as moisture

§18.2.4 Moisture-Evaporation Temperature 150

140
The adiabatic-saturation temperature combined with the wet-
bulb temperature can be used to track gas temperature, Tg, 130
and moisture-evaporation (solid) temperature, Ty, with 120 Gas

Temperature, °F
respect to location or time when removing surface moisture
110
from a wet solid with direct heat. The accuracy of the track-
ing is subject to the validity of the assumptions made. If the 100
moisture is water, Ty will be constant and equal to the con- Wet solid (at wet-bulb temperature)
90
stant, Tw, of the gas. If the moisture is an organic compound
80
with properties similar to those in Table 18.5, Ty will still be
equal to Tw, but, as shown next, will not be constant but will 70

decrease as gas temperature decreases. 60

Let To and be hypothetical entering conditions for a gas 50
used to dry a wet solid in an adiabatic, direct-heat dryer. At 0 1 2 3 4 5 6 7 8 9 10
Distance through dryer, ft
any point in the dryer, the conditions of the gas are given
(b) Toluene as moisture
from (18-24) as
Figure 18.21 Temperature profiles for dryer of Example 18.5.
DH vap
s
To Tg ¼ g o ð18-25Þ
ðC s Þo Since To is fixed, Ty must remain at the value Tw regard-
Assuming a quasi-steady-state transport condition at any less of the value of Tg. The result, for Example 18.5 below, is
point, the gas–liquid moisture interface conditions of Ty and shown in Figure 18.21 for air–water and air–toluene.
y are related by the following form of (18-18): For moisture other than water, (18-27) is appropriate.
ky M B DH vap Assume that the wet solid is dried batchwise, or flows cocur-
s
Ty Tg ¼ g y ð18-26Þ rently to the gas in a continuous dryer, with an initial temper-
h
ature equal to the wet-bulb temperature of the gas. Because
For a continuous dryer, (18-26) holds regardless of whether
the inverse of the psychrometric ratio is less than 1.0, Tw will
gas and wet solid flows are countercurrent or cocurrent.
decrease as Tg decreases, as in Example 18.5.
Equations (18-25) and (18-26) can be combined to give
Ty Tg ky M B ðCs Þo g y
¼ ð18-27Þ EXAMPLE 18.5 Temperature Variation in a Dryer.
To Tg h g o
Air enters a continuous, adiabatic direct-heat dryer at 140 F and 1
The coefficient of the RHS of (18-27) is the inverse of the atm with the relative humidity below, and exits at 100 F. The wet
psychrometric ratio. For air–water, it is 1:0, giving solid enters in cocurrent flow at the wet-bulb temperature of the
Ty Tg g y entering gas. Plot the variation of the moisture-evaporation tempera-
¼ ð18-28Þ ture, Ty, as a function of the distance through the dryer, z, for an
To Tg g o
exponential decrease of Tg according to the relation
Assume that the wet solid in contact with the initial or enter- Tg T s ¼ ðT o T s Þexpð 0:1377zÞ ð1Þ
ing gas is at the gas wet-bulb temperature, Tw. Then,
where z is in feet and temperatures are F, for: (a) water moisture
Tw Tg g w with entering air of 12.5% R , and (b) toluene moisture with enter-
¼ ð18-29Þ
To Tg g o ing air of 10% R .
748 Chapter 18 Drying of Solids

Solution Adiabatic-saturation
temperature line
Wet-bulb temperature
T o ¼ 140 F; T g in ¼ 140 F; and T g out ¼ 100 F lines ty
(a) From Figure 18.17, o ¼ 0:015 lb H2O/lb dry air. Tw ¼ 86.5 F = idi
um
h

Humidity
Ts. As the gas cools, its humidity follows the adiabatic-saturation on
r ati
line. Using (1) and Figure 18.17, tu
Sa

lb H2 O
z, ft Tg, F g;
lb dry air
0 140.0 0.015 o

2 127.1 0.018 Tv Tg Tv To
4 117.3 0.020 at at
6 109.9 0.022 Tg To

8 104.3 0.0235 Temperature

10 100.0 0.0245 Figure 18.22 Adiabatic drying path for general vapor–moisture
mixtures.
Equation (18-28) is satisfied only for Ty ¼ Tw ¼ 86.5 F and
y ¼ w ¼ 0:0275 lb H 2O/lb dry air
Tg, F Ty, F
For example, take Tg ¼ 109.9 F, g ¼ 0:022 lb H2O/lb dry air.
Using (18-28), take values of Ty ¼ 80 F and 90 F, and compute 140.0 92.0
the temperature and humidity ratios: 126.3 90.0
115.9 87.5
lb H2 O Ty Tg g y
Ty, F y; 107.9 86.5
lb dry air To Tg g o
100.0 85.5
86.5 0.0275 0.777 0.786
From (23), the following values of z are computed:
80.0 0.0223 0.993 0.043
90.0 0.0310 0.661 1.300
Tg, F z, ft
(b) From Figure 18.20 for To ¼ 140 F, R ¼ 10%: o ¼ 0:062 lb 140.0 0.0
toluene/lb dry air, Tw for entering air ¼ 92 F, and Ts ¼ 83 F. 126.3 2.3
From Table 18.4, 1/(psychrometric ratio) ¼ 1/1.908 ¼ 0.524. 115.9 4.5
From (18-27), 107.9 6.7
Ty Tg g y
100.0 9.6
¼ 0:524 ð2Þ
140 T g g 0:062
Thus, the moisture-evaporation temperature, plotted in Figure
From (1), T g ¼ T y þ ð140 T y Þexpð 0:1377zÞ ð3Þ 18.21b, decreases with decreasing gas temperature.
Equation (2) is solved iteratively for Ty for each value of
Tg, where y is the saturation humidity at Ty, as determined
from Figure 18.20. For example, for Tg ¼ 115.9 F, following the
adiabatic-saturation line, y ¼ 0:095, and (2) becomes:
§18.3 EQUILIBRIUM-MOISTURE CONTENT
T y 115:9 0:095 y OF SOLIDS
¼ 0:524
140 115:9 0:095 0:062
Faust et al. [9] group wet solids into two categories according
or T y ¼ 115:9 þ 382:7ð0:095 yÞ ð4Þ
to their drying behavior:
This equation is solved by assuming Ty, determining y, and
1. Granular or crystalline solids that hold moisture in
then computing Ty.
open pores between particles. These are mainly
from Fig. 18.20, lb
inorganic materials, examples of which are crushed
y
Assumed Ty, F toluene/lb dry air Ty, F from (4) rocks, sand, catalysts, titanium dioxide, zinc sulfate,
and sodium phosphates. During drying, the solid is
90 0.180 83.4 unaffected by moisture removal, so selection of drying
88 0.173 86.0 conditions and drying rate is not critical to the properties
87 0.165 89.0 and appearance of the dried product. Materials in this
category can be dried rapidly to very low moisture
By interpolation, Ty ¼ 87.5 F. contents.
This result can also be obtained graphically from Figure 18.20 2. Fibrous, amorphous, and gel-like materials that dis-
using the construction shown in Figure 18.22, with an essentially solve moisture or trap moisture in fibers or very fine
identical result. Calculations for the other values of Tg give
pores. These are mainly organic solids, including tree,
 §18.3 Equilibrium-Moisture Content of Solids 749

plant, vegetable, and animal materials such as wood, to expressions for humidity and is most convenient in drying
leather, soap, eggs, glues, cereals, starch, cotton, and calculations where the mass of bone-dry solid and dry gas
wool. These materials are affected by moisture remain constant while moisture is transferred from solid to
removal, often shrinking when dried and swelling gas. Less common is wt% moisture on a wet-solid basis, W.
when wetted. With these materials, drying in the later The two moisture contents are related by the expression
stages can be slow. If the surface is dried too rapidly,
moisture and temperature gradients can cause check- 100W
ing, warping, case hardening, and/or cracking. There- X¼ ð18-30Þ
100 W
fore, selection of drying conditions is a critical factor.
Drying to low moisture contents is possible only when
using a gas of low humidity. 100X
or W¼ ð18-31Þ
100 þ X
In a direct-heat drying process, the extent to which mois-
ture can be removed from a solid is limited, particularly for
Rarely used is moisture content on a volume basis because
the second category, by the equilibrium-moisture content of
wet solids of the second category shrink during drying. Also,
the solid, which depends on factors that include temperature,
moisture content is never expressed on a mole basis because
pressure, and moisture content of the gas. Even if the drying
the molecular weight of the dry solid may not be known.
conditions produce a completely dry solid, subsequent expo-
In Figure 18.23, equilibrium-moisture content, X , is plot-
sure of the solid to a different humidity can result in an
ted for a second-category solid for a given temperature and
increase in moisture content.
pressure, against relative humidity, R . In some cases, humid-
Terms used to describe equilibrium-moisture content are
ity, , is used with a limit of the saturation humidity, s . At
shown in Figure 18.23 with reference to a hypothetical equi-
R ¼ 100%, equilibrium-moisture content is called bound
librium isotherm. Moisture content, X, is expressed as mass
moisture, XB. If the wet solid has a total moisture content,
of moisture per 100 mass units of bone-dry solid. This is the
XT > XB, the excess, XT XB, is unbound moisture. At a rela-
most common way to express moisture content and is equiv-
tive humidity < 100%, the excess of XT over the equilibrium-
alent to wt% moisture on a dry-solid basis. This is analogous
moisture content, i.e., XT X , is the free-moisture content.
In the presence of a saturated gas, only unbound moisture
can be removed during drying. For a partially saturated gas,
only free moisture can be removed. But if R ¼ 0, all solids,
T, P = given enough time, may be dried to a bone-dry state. Solid
constants materials that can contain bound moisture are hygroscopic.
Bound moisture exhibits a vapor pressure less than the nor-
mal vapor pressure. The bound-moisture content of cellular
Total-moisture content, XT
materials such as wood is referred to as the fiber-saturation
Moisture content in solids, lb water/lb dry solids

point.
Free-moisture Unbound Experimental equilibrium-moisture isotherms at 25 C and
content, X = XT – X* moisture 1 atm are shown in Figure 18.24 for second-category materi-
als. At low values of R , e.g., <10%, moisture is bound to the
Bound-moisture content, XB solid on its surfaces as an adsorbed monomolecular layer.
Such bound moisture can also be present on solids of the first
category. At intermediate values of R , e.g., 20–60%, multi-
molecular layers may build up on the monolayer. At large
values of R , e.g., >60%, moisture is held in micropores so
small (e.g., <1 mm in radius) that vapor-pressure lowering
occurs, as predicted by the Kelvin equation (15-14). In cellu-
lar materials such as plant and tree matter, some moisture is
Bound held osmotically in fibers behind semipermeable membranes
moisture
of cell walls.
Temperature has a significant effect on equilibrium-
moisture content, an example of which is shown in Figure
18.25 for cotton at 96–302 F. At an R of 20%, equilibrium-
Equilibrium-moisture
content, X* moisture content decreases from 0.037 to 0.012 lb H2O/lb dry
cotton. Experimental determination of equilibrium-moisture
isotherms is complicated by a hysteresis effect, shown in
0
0 20 40 60 80 100 Figure 18.26 for sulfite pulp. Sorption and desorption curves
Percent relative humidity were obtained, respectively, by wetting and drying the solid,
Figure 18.23 Typical isotherm for equilibrium-moisture content of and it is seen that equilibrium-moisture content in drying is
a solid. always somewhat higher, particularly in the relative-humidity
750 Chapter 18 Drying of Solids

1.0

1— paper, newsprint 0.8

0.28
2— wool, worsted Adsorption

Relative saturation, pA/P sA

3— nitrocellulose
4— silk
5— leather, sole, oak tanned 0.6
0.24 6— kaolin
Moisture content in solids, lb water/lb dry solids

Desorption
7— N.C. tobacco leaf
8— soap
0.4
9— glue
10 — lumber
0.20
7 5 4
0.2

0.16 Desorption Adsorption

8 0
2 0 0.1 0.2 0.3
Equilibrium moisture content, lb water/lb dry solid
10
0.12 Figure 18.26 Effect of hysteresis on equilibrium-moisture content
of sulfite pulp.

3 9
0.08 range of 30% to 80%. According to Luikov [10], the hystere-
1
sis effect may be due to either: (1) failure to achieve true
equilibrium or (2) irreversibility of evaporation and conden-
0.04 sation in capillaries. For the latter, a possible explanation is
based on representations of moisture in necked capillaries, as
shown in Figure 18.26. For drying (desorption), the capillary
6
contains more moisture than for wetting (sorption). Thus, for
0
0 20 40 60 80 100 a given relative humidity, the equilibrium-moisture content
Percent relative humidity for drying is less than that for wetting.
Figure 18.24 Equilibrium-moisture content at 25 C and 1 atm. Bound moisture can also be defined as moisture held
chemically, as, for example, water of hydration of inorganic
crystals. This is one example of bound moisture dissolved in
a solid where the vapor pressure is lowered significantly
below the true vapor pressure. For inorganic salts that form
one or more hydrates, the hydrated form depends not only on
temperature, but also on relative humidity of the gas in con-
0.20 tact with the crystals. The effect of the latter for CuSO4,
in terms of the partial pressure of water, is shown in
Figure 18.27. At 25 C, the stable hydrate is CuSO4 5H2O.
96°F
However, if the partial pressure of H2O is 5.6–7.8 torr, the
0.16 113°F
Moisture content, lb H2O/lb dry solids

136°F
Vapor pressure of pure water
23.8
165°F
0.12 Unsaturated
Solution
Partial pressure of H2O, mm Hg

solutions
192°F Saturated
23.1
solution
Saturated solution
0.08 219°F +
CuSO4 • 5H2O CuSO4 • 5H2O

0.04 7.8
248°F CuSO4 • 3H2O
277°F 5.6
302°F CuSO4 CuSO4 • H2O
0 0.8
0 20 40 60 80 100
Relative humidity, percent 0 1 3 5 39.6
Equilibrium moisture content, moles H2O/mole CuSO4
Figure 18.25 Effect of temperature on the equilibrium-moisture
content of raw cotton at 1 atm. Figure 18.27 Equilibrium-moisture content for CuSO4 at 25 C.

trihydrate forms; at 0.8–5.6 torr, the monohydrate is favored.
Below 0.8 torr, CuSO4 crystals are free of water.
EXAMPLE 18.6 Effect of Equilibrium-Moisture
Content.
One-kg blocks of wet Borax laundry soap with an initial H2O con-
tent of 20.2 wt% on a dry basis are dried with air in a tunnel dryer at
1 atm. In the limit, if the soap were brought to equilibrium with the
air at 25 C and a relative humidity of 20%, determine the kg of
moisture evaporated from each block.
Solution
The initial moisture content of the soap on a wet basis is obtained
from a rearrangement of (18-30):
100X 100ð20:2Þ
W¼ ¼ ¼ 16:8 wt%
100 þ X 100 þ 20:2
Initial weight of moisture ¼ 0.168(1.0) ¼ 0.168 kg H2O
Initial weight of dry soap ¼ 1 0.168 ¼ 0.832 kg dry soap
From Figure 18.24, for soap at R ¼ 0:20, X* ¼ 0.037
Final weight of moisture ¼ 0.037(0.832) ¼ 0.031 kg
Moisture evaporated ¼ 0.168 0.031 ¼ 0.137 kg H2O/kg soap
§18.4 DRYING PERIODS
The decrease in average moisture content, X, as a function of
time, t, for drying either category of solids in a direct-heat
dryer was observed experimentally by Sherwood [11,12] to
exhibit the type of relationship shown in Figure 18.28a, pro-
vided the exposed surface of the solid is initially covered
with observable moisture. If that curve is differentiated with
respect to time and multiplied by the ratio of the mass of dry
solid to the interfacial area between the mass of wet solid and
the gas, a plot can be made of drying-rate flux, R,
dmy ms dX
R¼ ¼ ð18-32Þ
Adt A dt
where my ¼ mass of moisture evaporated and ms ¼ mass of
bone-dry solid as a function of moisture content, as shown in
Figure 18.28b. In Figure 18.28a, the final equilibrium-mois-
ture content is X . Although both plots exhibit four drying
periods, the periods are more distinct in the drying-rate
curve. For some wet materials and/or some hot-gas condi-
tions, fewer than four drying periods are observed. From A to
B, the wet solid is being preheated to an exposed-surface
temperature equal to the wet-bulb gas temperature, while
moisture is evaporated at an increasing rate. At the end of the
preheat period, if the wet solid is of the granular, first cate-
gory, a cross section has the appearance of Figure 18.29a,
where the exposed surface is still covered by a film of mois-
ture. A wet solid of the second category is covered on the
exposed surface by free moisture. The drying rate now
§18.4 Drying Periods 751
Moisture content, mass moisture/mass dry solid
A
B
Critical-
moisture C
content
D
Equilibrium-moisture content, X*
Time
(a) Moisture content
Critical-moisture content
Rate of drying, mass evaporated/
C B
time-unit surface area
A
D
Moisture content, mass moisture/mass dry solid
(b) Drying rate
Figure 18.28 Drying curves for constant drying conditions.
becomes constant during the period from B to C, which pre-
vails as long as free moisture covers the exposed surface.
This surface moisture may be part of the original moisture
that covered the surface, or it may be moisture brought to the
surface by capillary action in the case of wet solids of the
first category or by liquid diffusion in the case of wet solids
of the second category. In either case, the rate of drying is
controlled by external mass and heat transfer between the
exposed surface of the wet solid and the bulk gas. Migration
of moisture from the interior of the wet solid to the exposed
surface is not a rate-affecting factor. This period, the con-
stant-rate drying period, terminates at point C, the critical
moisture content. When drying wet solids of the first cate-
gory under agitated conditions—as in a direct-heat rotary
dryer, fluidized-bed dryer, flash dryer, or agitated batch
dryer—such that all particle surfaces are in direct contact
with the gas, the constant-rate drying period may extend all
the way to X .
At C, the moisture just barely covers the exposed surface;
and then until point D is reached, as shown in Figure 18.29b,
the surface tends to a dry state because the rate of liquid
travel by diffusion or capillary action to the exposed surface
is not sufficiently fast. In this period, the exposed-surface
temperature remains at the wet-bulb temperature if heat con-
duction is adequate, but the wetted exposed area for mass
transfer decreases. Consequently, the rate of drying decreases
752 Chapter 18 Drying of Solids
Drying-gas flow
(a) Constant-rate period
(b) First falling-rate period
(c) Second falling-rate period
Figure 18.29 Drying stages for granular solids.
linearly with decreasing average moisture content. This is the
first falling-rate drying period. It is not always observed with
wet solids of the second category.
During the period from C to D, the liquid in the pores of
wet solids of the first category begins to recede from the
exposed surface. In the final period from D to E, as shown in
Figure 18.29c, evaporation occurs from liquid surfaces in the
pores, where the wet-bulb temperature prevails. However, the
temperature of the exposed surface of the solid rises to
approach the dry-bulb temperature of the gas. During this
period, the second falling-rate drying period, the rate of dry-
ing may be controlled by vapor diffusion for wet solids of the
first category and by liquid diffusion for wet solids of the sec-
ond category. The rate falls exponentially with decreasing
moisture content.
§18.4.1 Constant-Rate Drying Period
In direct-heat equipment, drying involves transfer of heat
from the gas to the surface and interior of the wet solid, and
mass transfer of moisture from the interior and surface of
the solid to the gas. During the constant-rate period, the
rate of mass transfer is determined by gas-phase bound-
ary-layer or film resistance at the wet surface of the solid.
The wet solid is assumed to be at a uniform temperature,
so the only resistance to convective heat transfer is in the
gas phase. The rate of moisture evaporation can then be
based on convective heat transfer or mass transfer,
according to conventional, but simplified, transport rela-
tionships in which thermal radiation, the bulk-flow effect
for mass transfer, and the sensible-heat effect for the
evaporated moisture are often ignored:
dmy h T g T i A
¼ ¼ M A ky yi yg A ð18-33Þ
dt DH vap
i
where subscript i refers to the gas–solid interface.
As discussed previously, the interface at these conditions
is at the wet-bulb temperature, Tw. Although drying-rate cal-
culations could be based on mass transfer using (18-33), it is
more common to use the heat-transfer relation of (18-33)
when air is the gas and water is the moisture because of the
wide availability of the psychrometric chart for that system,
the equality of wet-bulb and adiabatic-saturation tempera-
tures, and a wider availability of correlations for convective
heat transfer than for mass transfer, although analogies can
be used to derive one from the other. Combining (18-32) and
(18-33), the drying-rate flux for constant-rate drying period
Rc, in terms of heat transfer, becomes
h Tg Tw
Rc ¼ ð18-34Þ
DH vap
w
while an equivalent, but less-useful, mass-transfer form is
obtained by combining (18-16), (18-32), and (18-33):
Rc ¼ M B k y ð w dÞ ð18-35Þ
where subscripts d and w refer to gas dry-bulb and wet-bulb
conditions.
For some dryers, it is preferable to use a volumetric heat-
transfer coefficient, (ha), defined by
dmy ðhaÞ T g T i V
¼ ð18-36Þ
dt DH vap
w
where a ¼ external surface area of wet solids per unit volume
of dryer, and V ¼ volume of dryer. Then, the drying rate per
unit dryer volume during the constant-rate drying period is
ðhaÞðT d T w Þ
ðRc ÞV ¼ ð18-37Þ
DH vap
w
Interphase heat-transfer coefficients were discussed for
several geometries in §3.5. Empirical equations useful for
drying-rate calculations are summarized in Mujumdar [1],
and representative equations, when the gas is air, are listed in
Table 18.6, where in (1), G is the mass velocity of air in the
flow channel that passes over the wet surface. In (2), G is the
mass velocity of the air impinging on the wet surface. In (3)
to (8), dp is the particle diameter and G is the superficial mass
velocity.
 §18.4 Drying Periods 753

Table 18.6 Empirical Equations for Interphase Heat-Transfer For the constant-rate drying period, the heat-transfer form of (18-33)
Coefficients for Application to Dryers (h in W/m2-K, G in kg/hr- applies, which upon integration gives
m2, dp in m)
my DH vap
w
tc ¼
Geometry Equation hðT d T w ÞA
Flat-plate, parallel flow h ¼ 0.0204G0.8 where tc is the time to reach the critical moisture content. From the
(Td ¼ 45 150 C, humidity chart of Figure 18.17,
G ¼ 2,450 29,300) (1) lb H2 O
Flat-plate, perpendicular, h ¼ 1.17G0.37 T w ¼ 100 F and ¼ 0:026
lb dry air
impingement flow (G ¼ 3,900 19,500) (2)
Td ¼ 170 F ¼ 76.7 C and T d T w ¼ 170 100 ¼ 70 F
Packed beds, h ¼ 0:151G 0:59
=d 0:41
p ; ¼ 38:9 K
through-circulation ðN Re > 350Þ (3)
At T w ¼ 100 F; DH vap
w ¼ 1037:2 Btu=lb ¼ 2; 413 kJ=kg
h ¼ 0:214G0:49 =d 0:51
p ;
From (18-10), Table 18.4,
ðN Re < 350Þ (4)
1 0:026
Fluidized beds N Nu ¼ 0:0133N 1:6
Re yH ¼ 0:730ð170 þ 460Þ þ ¼ 16:5 ft 3 =lb dry air
(5) 28:97 18:02
ð0 < N Re < 80Þ
Pneumatic conveyors N Nu ¼ 0:316N 0:8 or,
Re
ð8 < N Re < 500Þ (6) 16:5=ð1 þ 0:026Þ ¼ 16:1 ft3 =lb moist air ¼ 1:004 m3 =kg moist air
Droplets in spray dryers N Nu ¼ 2 þ 1:05N 0:5
1=3 0:175 ¼ 1=r
Re N Pr N Gu
ðN Re < 1000Þ (7) G ¼ uavg r ¼ uavg =y ¼ 4=1:004 ¼ 3:98 kg=m2 -s ¼ 14; 300 kg=m2 -h

N Nu ¼ 0:0005N 1:46 1=3 and A ¼ 1:5 m2

Spouted beds Res ðu=us Þ (8)
From Table 18.6, (1) applies for turbulent flow with Td and G within
N Re ¼ d p G=m; N Nu ¼ hd p =k; N Pr ¼ CP m=k; N Res ¼ d p Gs =m the allowable range.
Gs ¼ mass velocity for incipient spouting
u ¼ velocity, us ¼ incipient spouting velocity h ¼ 0:0204ð14; 300Þ0:8 ¼ 43 W=m2 -K ¼ 43 J=s-m2 -K
N Gu ¼ ðT d T w Þ=T d in absolute temperature
dp ¼ particle size, CP ¼ specific heat of gas
From (1), using SI units
m ¼ viscosity of gas, k ¼ thermal conductivity of gas
ð11:2Þ½ð2; 413Þð1; 000Þ
tc ¼ ¼ 10; 800 s ¼ 2:99 h
ð43Þð38:9Þð1:5Þ

The dramatic effect of exposed surface area of wet solids

in drying was shown by Marshall and Hougen [13] and is
illustrated in the next two examples, which deal with batch
drying. In Example 18.7, cross-circulation, batch tray drying
is used to dry slabs of filter cake. In Example 18.8, the filter
cake is extruded and then dried by through-circulation. The
difference in the two drying times for the constant-rate dry-
ing period is found to be very significant.
EXAMPLE 18.7 Batch Drying with Cross-Circulation.
CaCO3 filter cake in a tray is to be dried by cross-circulation from the
top surface. Each tray is 2.5 cm high, with an area of 1.5 m2, and is
filled with 73 kg of wet filter cake having a water content of 30% on
the dry basis. The heating medium is air at 1 atm and 170 F with a
relative humidity of 10%. The velocity of air passing across the wet
solid is 4 m/s. Estimate time in hours needed to reach the experimen-
tally determined, critical moisture content (end of the constant-rate
period) of 10% on the dry basis, if the preheat period is neglected.
EXAMPLE 18.8 Batch Drying with
Through-Circulation.
The filter cake of Example 18.7 is extruded into cylindrical-shaped
pieces of 1/4-inch diameter and 1/2-inch length to form a bed 1.5 m2
in cross-sectional area and 5 cm high, with an external porosity of
50%. Air at 170 F and 10% relative humidity passes through the
bed at a superficial velocity of 2 m/s (average interstitial velocity of
4 m/s). Estimate the time in hours needed to reach the critical-mois-
ture content, if the preheat period is neglected. Compare this time to
that estimated in Example 18.7.
Solution
Compared to the tray of Example 18.7, the bed is twice as high
with the same cross-sectional area. Therefore, for a porosity of
50%, the bed contains the same amount of wet solids. Thus, as in
Example 18.7,
mwet cake ¼ 73 kg; my ¼ 11:2 kg H2 O evaporated; DH vap
w ¼ 2; 413 kJ=kg;

Solution and T d T w ¼ 38:9 K

30 Assume the extrusion density equals filter-cake density.

H2 O in wet cake ¼ ð73Þ ¼ 16:8 kg;
130 73
r filter cake ¼ ¼ 1; 950 kg=m3
2:5
H2O in cake at X c ¼ 0:10ð73 16:8Þ ¼ 5:6 kg; 1:5
my ¼ H2 O evaporated ¼ 16:8 5:6 ¼ 11:2 kg 100
754 Chapter 18 Drying of Solids

However, in Example 18.7, the moisture may have to

3:14ð0:25Þ2 ð0:5Þ
Volume of one extrusion ¼ pD2 L=4 ¼ travel from as far away as 25 mm to reach the exposed sur-
4
¼ 0:0245 in3 ¼ 4:01 10 7 m3
face, while in Example 18.8, the distance is only 3.2 mm.
Therefore, as a first approximation, it might be expected that
2:5 7
Number of extrusions ¼ 1:5 ð0:5Þ=4:01 10 the critical moisture contents for the two examples might not
100
be the same. The value of 10% on the dry basis was taken
¼ 46; 800 from through-circulation drying experiments.
Surface area=extrusion ¼ pDL þ pD2 =2 When moisture travels from the interior of a wet solid to
" #
ð0:25Þ2 the surface, a moisture profile develops in the wet solid. The
¼ 3:14 ð0:25Þð0:50Þ þ profile’s shape depends on the nature of the moisture move-
2
ment, as discussed by Hougen, McCauley, and Marshall [16].
¼ 0:49 in:2 ¼ 0:000316 m2
If the wet solid is of the first category, where the moisture is
A ¼ 46; 800ð0:000316Þ ¼ 14:8 m2 not held in solution or in fibers but is free moisture in
Thus, the transport area is 14.8/1.5 ¼ 9.9 times that for Example the interstices of particles like soil and sand, or is moisture
18.7. From Table 18.6, (3) or (4) applies for estimating h, depending above the fiber-saturation point in paper and wood, then
on NRe. From Example 18.7, but with a superficial bed velocity of moisture movement occurs by capillary action. For wet solids
50% of the crossflow velocity, G ¼ 3.98/2 ¼ 1.96 kg/m2-s. of the second category, the internal moisture is bound mois-
Equations (3) and (4) refer to the work of Gamson, Thodos, and ture, as in the last stages of drying of paper and wood, or
Hougen [14] for N Re > 350 and Wilke and Hougen [15] for soluble moisture, as in soap and gelatin. This type of mois-
N Re < 350, respectively. For both correlations, dp is taken as the ture migrates to the surface by liquid diffusion. Moisture can
diameter of a sphere of the same surface area as the particle. For also migrate by gravity, external pressure, and by vaporiza-
the extrusions of this example with L ¼ 2D,
tion–condensation sequences in the presence of temperature
pD2 gradients. In addition, vapor diffusion through solids can
pd 2p ¼ þ 2pD2 ¼ 2:5pD2
2 occur in indirect-heat dryers when heating and vaporization
Solving (1), occur at opposed surfaces.
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi A moisture profile for capillary flow is shown in
d p ¼ D 2:5 ¼ 0:25 2:5 ¼ 0:395 in: ¼ 0:010 m
Figure 18.30a. It is concave upward near the exposed surface,
m 0:02 cP ¼ 2 10 5 kg=m-s concave downward near the opposed surface, with a point of
d p G ð0:010Þð1:96Þ inflection in between. For flow of moisture by diffusion, as in
N Re ¼ ¼ ¼ 980
m 2 10 5 Figure 18.30b, the profile is concave downward throughout.
Therefore, (3) applies and If the diffusivity is independent of moisture content, the solid
J curve applies. If, as is often the case, the diffusivity decreases
h ¼ 0:151ð14; 300=2Þ0:59 =ð0:010Þ0:41 ¼ 188 with moisture content, due mainly to shrinkage, the dashed
s-m2 -K
profile applies.
The h is 188/43 ¼ 4.4 times greater than in Example 18.7. From
During the falling-rate period, idealized theories for capil-
(1) in that example,
lary flow and diffusion can be used to estimate drying rates.
ð11:2Þ½ð2; 413Þð1; 000Þ Alternatively, estimates could be made by a strictly empirical
tc ¼ ¼ 250 s ¼ 4:16 min
ð188Þð38:9Þð14:8Þ approach that ignores the mechanism of moisture movement,
This, and the preceding, example show that cross-circulation drying but relies on experimental determination of drying rate as a
takes hours, whereas through-circulation drying may require only function of average moisture content for a particular set of
minutes. conditions.

Empirical Approach
§18.4.2 Falling-Rate Drying Period The empirical approach relies on experimental data in the
When the drying rate in the constant-rate period is high and/ form of Figure 18.31a (Case 1), and Figures 18.31b (Case 2)
or the distance that interior moisture must travel to reach the and 18.31c (Case 3), where, for all cases, the preheat period
surface is large, moisture may fail to reach the surface fast is ignored. In these plots, the abscissa is the free-moisture
enough to maintain a constant drying rate, and a transition to content, X ¼ XT X, shown in Figure 18.23, which allows all
the falling-rate period occurs. In Examples 18.7 and 18.8, the three plots to be extended to the origin, if all free moisture is
constant drying rates from (18-34) are removed.
From (18-32),
43ð38:9Þð3; 600Þ Z Z
Rc ¼ ¼ 2:50 kg=h-m 2 ms dX
ð2; 413Þð1; 000Þ dt ¼ ð18-38Þ
A R
ð188Þð38:9Þð3; 600Þ Ignoring preheat, for the constant-rate period, R ¼ Rc ¼
and Rc ¼ ¼ 10:9 kg=h-m2 constant. Starting from an initial free-moisture content of Xo
ð2; 413Þð1; 000Þ
at time t ¼ 0, the time to reach the critical free-moisture
 §18.4 Drying Periods 755

Drying rate
Moisture content

Center line of wet solid

Free-moisture content
(a) Empirical case 1

Distance from surface

Drying rate
(a) Moisture flow by capillary action
Moisture content

Center line of wet solid

Actual

Theoretical Free-moisture content

(b) Empirical case 2
Drying rate

Distance from surface

(b) Moisture flow by diffusion
Figure 18.30 Moisture distribution in wet solids during drying.
[From W.L. McCabe, J.C. Smith, and P. Harriott, Unit Operations of Chemi-
cal Engineering, 5th ed., McGraw-Hill, New York (1993) with permission.]

content, Xc, at time t ¼ tc is obtained by integrating (18-38): Free-moisture content

(c) Empirical case 3
ms ðX o X c Þ Figure 18.31 Drying-rate curves.
tc ¼ ð18-39Þ
ARc
For Case 1 (Figure 18.31a) of the falling-rate period, the
rate of drying is linear with X and terminates at the origin,
EXAMPLE 18.9 Constant- and Falling-Rate Periods.
according to
R ¼ Rc X=X c ð18-40Þ Marshall and Hougen [13] present experimental data for the
through-circulation drying of 5/16-inch extrusions of ZnO in a bed
Substituting (18-40) into (18-39) and integrating t from tc of 1 ft2 cross section by 1-inch high, using air of Td ¼ 158 F and
to t > 0 and X from Xc to X > 0 gives the following expres- Tw ¼ 100 F at a flow rate of 340 ft3/min. The data show a constant-
sion for the drying time in the falling-rate period, tf: rate period from Xo ¼ 33% to Xc ¼ 13%, with a drying rate of 1.42 lb
H2O/h-lb bone-dry solid, followed by a falling-rate period that
ms X c Xc ms X c Rc approximates Case 1 in Figure 18.31a. Calculate the drying time for
tf ¼ t tc ¼ ln ¼ ln ð18-41Þ the constant-rate period and the additional time in the falling-rate
ARc X ARc R
period to reach a free-moisture content, X, of 1%.
The total drying time, t T , is the sum of (18-39) and
(18-41): Solution
ms Xc In (18-38), the drying rate, R, corresponds to mass of moisture
tT ¼ t c þ tf ¼ ðX o X c Þ þ X c ln ð18-42Þ evaporated per unit time per unit of exposed area of wet material. In
ARc X
756 Chapter 18 Drying of Solids

this example, drying rate is not given per unit area, but per mass with the constraints that R ¼ Rc1 at X c1 , and that R ¼ Rc2 at X c2 . In
of bone-dry solid, with some associated exposed area. Equations the second subregion, (18-44) applies, but with the constraint that
(18-38) and (18-42) are rewritten in terms of R0 ¼ RA/ms as R ¼ Rc2 at X c2 .
Z Z
dX
dt ¼ 0 ð1Þ
R
1 Xc
and tT ¼ tc þ tf ¼ 0 ðX o X c Þ þ X c ln ð2Þ
Rc X EXAMPLE 18.11 Complex Falling-Rate Period.
From (2), for just the constant-rate period, Experimental data of Sherwood [12] for the surface drying of a
1 3.18-cm-thick 6.6-cm2 cross-sectional area slab of a thick paste
tc ¼ ½0:33 0:13 ¼ 0:141 h ¼ 8:45 min
1:42 of CaCO3 (whiting) from both sides by air at Td ¼ 39.8 C and
From (2), for just the falling-rate period, Tw ¼ 23.5 C and a cross-circulation velocity of 1 m/s exhibit the
complex type of drying-rate curve shown in Figure 18.31c, with the
1 0:13 following constants:
tf ¼ 0:13 ln ¼ 0:235 h ¼ 14:09 min
1:42 0:01
Constant-rate period:
The total drying time, ignoring the preheat period, is
X o ¼ 10:8%; X c1 ¼ 8:3%; and Rc1 ¼ 0:053 g H2 O=h-cm2
tT ¼ tc þ tf ¼ 8:45 þ 14:09 ¼ 22:5 min
First falling-rate period:
For Case 2 of Figure 18.31b, R in the falling-rate period can be
expressed as a parabolic function. X c2 ¼ 3:7% and Rc2 ¼ 0:038 g H2 O=h-cm2
R ¼ aX þ bX 2 ð18-43Þ Second falling-rate period to X ¼ 2.2%:
The values of the parameters a and b are obtained by fitting (18-43) R ¼ 29:03 X 2 0:048 X ð1Þ
to the experimental drying-rate plot, subject to the constraint that
Determine the time to dry a slab of the same dimensions at the same
R ¼ Rc at X ¼ Xc. If (18-43) is substituted into (18-38) and the result
drying conditions, but from Xo ¼ 0.14 to X ¼ 0.01, ignoring the pre-
is integrated for the falling-rate period from X ¼ Xc to some final
heat period. Assume an initial weight of 46.4 g.
value of Xf and corresponding Rf, the time for the falling-rate period
is found to be
" # Solution
ms X c a þ bXf ms X 2c Rf
tf ¼ t tc ¼ ln ¼ ln 2 ð18-44Þ Constant-rate period:
aA X f ða þ bX c Þ aA X f Rc
X o ¼ 0:14; X c1 ¼ 0:083; and Rc1 ¼ 0:053 g=h-cm2

1
ms ¼ 46:4 ¼ 40:7 g of moisture-free solid
1:14
A ¼ 2ð6:6Þ ¼ 13:2 cm2 ðdrying is from both sidesÞ
EXAMPLE 18.10 Falling-Rate Period by Empirical
Equation. 40:7ð0:14 0:083Þ
From (18-39), tc ¼ ¼ 3:32 h
Experimental data for through-circulation drying of 1/4-inch- 13:2ð0:053Þ
diameter spherical pellets of a nonhygroscopic carburizing com- First falling-rate period:
pound exhibit constant-rate drying of 1.9 lb H2O/h-lb dry solid from In this period, R is linear with end points (Rc1 ¼ 0:053, X c1 ¼ 0:083)
Xo ¼ 30% to Xc ¼ 21%, followed by a falling-rate period to Xf ¼ 4% and (Rc2 ¼ 0:038, X c2 ¼ 0:037). This gives for (18-45),
that fits (18-43) with a ¼ 3.23 and b ¼ 27.7 (both in lb H2O/h-lb dry
R ¼ 0:0259 þ 0:326X ð2Þ
solid) for X as a fraction and R replaced by R0 in lb H 2O/h-lb dry
solid. Calculate the time for drying in the falling-rate period. Note Substituting (2) into (18-38) and integrating,
0 Z
that the values of a and b satisfy the constraint of R c ¼ 1:9 at ms xc2 dX
tf 1 ¼
Xc ¼ 0.21. A xc1 0:0259 þ 0:326X
ms 1 0:0259 þ 0:326X c1
Solution ¼ ln
A 0:326 0:0259 þ 0:326X c2
For R0 in the given units, (18-44) becomes ms 1 Rc1
¼ ln
A 0:326 Rc2
1 X c a þ bX f
tf ¼ ln ð1Þ 40:7 0:053
a X f ða þ bX c Þ ¼ ln ¼ 3:15 h
Thus, 13:2ð0:326Þ 0:038
Second falling-rate period:
1 0:21 3:23 þ 27:7ð0:04Þ
tf ¼ ln ¼ 0:286 h ¼ 17:1 min This period extends from X c2 ¼ 0:037 to X ¼ 0.022, with R given by
3:23 0:04 3:23 þ 27:7ð0:21Þ (1) for (18-43), with a ¼ 0.048 and b ¼ 29.03. From (18-44),
For Case 3 of Figure 18.31c, the falling-rate period consists of 40:7 0:037½ 0:048 þ 29:03ð0:022Þ
tf 2 ¼ ln ¼ 2:08 h
two subregions. In the first subregion, which is linear, ð 0:048Þð13:2Þ 0:022½ 0:048 þ 29:03ð0:037Þ
R ¼ aX þ b ð18-45Þ and the total drying time is tT ¼ 3:32 þ 3:15 þ 2:08 ¼ 8:6 h
 §18.4 Drying Periods 757

For drying-rate curves of shapes other than those of Figure Liquid-Diffusion Theory
18.31, time for drying from any Xo to any X can be deter-
The empirical approach for determining drying time in the
mined by numerical or graphical integration of (18-38) or (1)
falling-rate period is limited to the conditions for which the
in Example 18.9, as illustrated in the following example.
experimental drying-rate curve is established. A more gen-
eral approach, particularly for nonporous wet solids of the
second category, is the use of Fick’s laws of diffusion. Once
EXAMPLE 18.12 Drying Time from Data. the diffusion coefficient is established from experimental
Marshall and Hougen [13] present the following experimental data data for a wet solid, Fick’s laws can be used to predict drying
for the through-circulation drying of rayon waste. Determine the rates and moisture profiles for wet solids of other sizes and
drying time if X0 ¼ 100% and the final X is 10%. Assume that all shapes and drying conditions during the falling-rate period.
moisture is free moisture. Mathematical formulations of liquid diffusion in solids are
readily obtained by analogy to the solutions available for tran-
lb H2 O sient heat conduction in solids, as summarized, for example,
R0 ; by Carslaw and Jaeger [17] and discussed in §3.3. Two solu-
X, lb H2O/lb dry solid h-lb dry solid
tions are of particular interest for drying of slabs in the falling-
1.40 24 rate period, where the area of the edges is small compared to
1.00 24 the area of the two faces, or the edges are sealed to prevent
0.75 24 escape of moisture. As in heat-conduction calculations, the
0.73 21 equations also apply when one of the two faces is sealed. Two
0.70 18 general moisture-distribution cases are considered.
0.65 15.3
0.55 13 Case 1: Initially uniform moisture profile in the wet solid
0.475 12.3 with negligible resistance to mass transfer in the gas
0.44 12.2 phase.
0.40 11
Case 2: Initially parabolic moisture profile in the wet solid
0.20 5.5
with negligible resistance to mass transfer in the gas
0 0
phase.
Although the equations for these two cases are developed
Solution here only for a slab with sealed edges, other solutions are
available in Carslaw and Jaeger [17]. When edges of slabs
The data are plotted in Figure 18.32, where three distinct drying-rate and cylinders are not sealed, Newman’s method [18], as
periods are seen, but the two falling-rate periods are in the reverse discussed in §3.3, is suitable.
order of Figure 18.31c. By numerical integration of Equation (1) in Case 1. This case models slow-drying materials for which
Example 18.9 with a spreadsheet, the following drying times are
the rate of drying is controlled by internal diffusion of mois-
obtained, noting that R0 ¼ 2.75 lb H2O/h-lb dry solid at X ¼ 0.10,
Rc1 ¼ 24 at X c1 ¼ 0:75, and Rc2 ¼ 12:2 at X c2 ¼ 0:44.
ture to the exposed surface. This occurs if, initially, the wet
solid has no surface liquid film and external resistance to
tc ¼ 0:027 h ¼ 1:63 minutes; tf 1 ¼ 1:28 minutes; and tf 2 ¼ 3:21 minutes
mass transfer is negligible, thus eliminating the constant-rate
tT ¼ 1:63 þ 1:28 þ 3:21 ¼ 6:12 minutes drying period. Alternatively, the wet solid can have a surface
liquid film, but during the evaporation of that film in a
25 constant-rate drying period controlled by gas-phase mass
transfer, no moisture diffuses to the surface, and after com-
Rate of drying, lb water/h-lb dry solid

pletion of evaporation of that film, resistance to mass transfer

20
is due to internal diffusion in a falling-rate period.
The slab, of thickness 2a, is pictured in Figure 3-7a, where
15 the edges at x ¼ c and y ¼ b are sealed to mass transfer.
Internal diffusion of moisture is in the z-direction only
toward exposed faces at z ¼ a. Alternatively, the slab may
10
be of thickness a with the face at z ¼ 0 sealed to mass trans-
fer. Initially, the moisture content throughout the slab, not
5 counting any surface liquid film, is assumed uniform at Xo.
At the beginning of the falling-rate period, t ¼ 0, the exposed
face(s) is(are) brought to the equilibrium-moisture content,
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 X . For constant moisture diffusivity, DAB, Fick’s second
Moisture content, lb water/lb dry solid law, as discussed in §3.3, applies:
Figure 18.32 Data for through-circulation drying of rayon waste. @X @2X
¼ DAB 2 ð18-46Þ
@t @z
758 Chapter 18 Drying of Solids

for t 0 in the region a z a, where the boundary con- Included are values of Eavg, computed from its definition in (18-49),
ditions are X ¼ Xo at t ¼ 0 for a < z < a and X ¼ X at z ¼ and values of t/a2, where a ¼ 0.5 (1.90) ¼ 0.95 cm.
a for t 0.
The solution to (18-46) for the moisture profile as a func- t, h Xavg, g H2O/g dry wood t/a2, h/cm2 Eavg
tion of time under these boundary conditions, as discussed in
0.36 0.362 0.40 0.900
§3.3 and first proposed for drying applications by Sherwood
0.90 0.328 1.00 0.800
[11], is in terms of the unaccomplished free-moisture change,
1.53 0.303 1.70 0.730
and a modification of (3-80) applies:
1.94 0.291 2.15 0.694
X X 4X 1
ð 1Þn 2.89 0.267 3.20 0.626
E¼ ¼
Xo X p n¼0 ð2n þ 1Þ 3.47 0.255 3.85 0.591
" # 4.02 0.245 4.45 0.562
p2 ð2n þ 1Þ2 DAB t pð2n þ 1Þ z 4.92 0.230 5.45 0.520
exp 2
cos
4 a 2 a 5.82 0.218 6.45 0.483
6.95 0.204 7.70 0.443
ð18-47Þ
8.03 0.192 8.90 0.409
Thus, E is a function of two dimensionless groups, the Fou- 8.98 0.183 9.95 0.382
rier number for diffusion, N FoM ¼ DAB t=a2 , and the position
ratio, z/a. This solution is plotted as (1 E) in terms of these Using the data, determine the average value of the diffusivity by
two groups in Figure 3.8. nonlinear regression of (18-49), and use that value to determine the
The rate of mass transfer from one face is given by (3-82), drying time from Xo ¼ 45% to X ¼ 10% with X ¼ 6% for a piece of
which in terms of R, the drying rate in mass of moisture poplar measuring 72 inches long 12 inches wide 1 inch thick,
evaporated per unit time per unit area, is neglecting mass transfer from the edges and assuming only a fall-
ing-rate period, with negligible resistance in the gas phase.
2DAB ðX o X Þrs
R¼
a Solution
" # ð18-48Þ
X 1
p2 ð2n þ 1Þ2 DAB t 1
exp ms ¼ 264 ¼ 189 g dry wood
n¼0
4 a2 1 þ 0:397

where rs ¼ mass of dry solid/volume of slab. A for two faces ¼ 2(15.2)2 ¼ 462 cm2
Also of interest is the average moisture content of the slab At any instant, from (18-38),
during drying. From (3-85), ms dX avg
R¼ ð1Þ
A dt
X avg X
Eavg ¼ From a plot of the data, approximate values of R as a function of
Xo X
" # Xavg are computed to be
8 X 1
1 p2 ð2n þ 1Þ2 DAB t
¼ 2 exp
p n¼0 ð2n þ 1Þ2 4 a2 R, g H2O/h-cm2 Xavg, g H2O/g dry solid
ð18-49Þ 0.02622 0.345
Equations (18-47)–(18-49) can be used to determine the 0.01573 0.315
moisture diffusivity, DAB, from experimental data, and then 0.01258 0.297
that value can be used to estimate drying rates for other con- 0.01019 0.279
ditions, as illustrated in the next example. However, such cal- 0.00847 0.261
culations must be made with caution because often the 0.00760 0.250
diffusivity is not constant, as shown by Sherwood [11] for 0.00661 0.238
drying of slabs of soap, but decreases with decreasing mois- 0.00582 0.224
ture content due to of shrinkage and/or case hardening. In 0.00503 0.211
that case, numerical solutions are necessary. 0.00446 0.198
0.00404 0.187

EXAMPLE 18.13 Drying by Liquid Diffusion.
A piece of poplar wood 15.2 cm long 15.2 cm wide 1.9 cm
thick, with the edges sealed with a waterproofing cement, was dried
from both faces in a tunnel dryer using cross-circulation of air at
1 m/s. Initial moisture content was 39.7% on the dry basis, initial
weight of the wet piece was 264 g, and no shrinkage occurred during
drying. The direction of diffusion was perpendicular to the grain.
The equilibrium-moisture content was 5% on the dry basis. Data
were obtained for the moisture content as a function of time.
These results are plotted in Figure 18.33, where it appears that all of
the drying takes place in the falling-rate period. Thus, the data may
be consistent with the Case 1 diffusion theory.
To determine the average moisture diffusivity, a spreadsheet is
used to prepare a semilog plot of the data points as Eavg against t/a2,
as shown in Figure 18.34. Equation (18-49) is then evaluated on the
spreadsheet for different values of the moisture diffusivity until the
best fit of the data is obtained, based on minimizing the error sum of
squares (ESS) of the differences between Eavg of the data points and
the corresponding Eavg values calculated from (18-49).
 §18.4 Drying Periods 759

0.030 Since N FoM > 0:1, (2) and (3) are valid, and
a ¼ 0:5 in: ¼ 1:27 cm
0.025
Rate of drying, g water/h-cm2

6
DAB ¼ 9:0 10 cm2 =s from the above experiments
0.020 a2 N FoM ð1:27Þ2 ð0:839Þ
t ¼ ¼ ¼ 41:8 h
DAB 9:0 10 6 ð3600Þ
0.015

0.010 Case 2. When a liquid-diffusion-controlled, falling-rate

drying period is preceded by a constant-rate period, that rate
0.005 of drying is determined by external mass transfer in the gas
phase, as discussed earlier, but diffusional resistance to the
0.000 flow of moisture in the solid causes a parabolic moisture pro-
0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 file to be established in the solid, as discussed by Sherwood
Average moisture content, g water/g dry wood
[19] and Gilliland and Sherwood [20].
Figure 18.33 Experimental data for drying poplar wood. For the slab of Figure 3.7a, Fick’s second law, as given by
(18-46), still applies, with X ¼ Xo at t ¼ 0 for a < z < a.
1.0 However, during the constant-rate drying period, the slab–gas
0.8
interface boundary conditions are changed from those of Case
1 to the conditions @X/@z ¼ 0 at z ¼ 0 for t 0 and Rc ¼
0.6 DAB rs ð@X=@zÞ at z ¼ a for t 0. This latter boundary
Eavg = (Xavg – X*)/(Xo – X*)

condition is more conveniently expressed in the form

0.4 @X Rc
¼ ð18-50Þ
@z rs DAB
where the term on the RHS is a constant during the constant-
rate period. This is analogous to a constant-heat-flux bound-
0.2
ary condition in heat transfer. The solution for the moisture
Experimental data profile as a function of time during the constant-rate drying
Theory, (18-49) with DAB = 9.0 × 10–6 cm2/s period is given by Walker et al. [21] as:
0.1 Rc a 1 z 2 1 DAB t
0 2 4 6 8 10 X ¼ Xo þ 2
2
t/a h/cm 2 DAB rs 2 a 6 a
)
Figure 18.34 Best fit by diffusion theory of experimental data for 2 X1
ð 1Þm 2 2 DAB t pmz
drying poplar wood. exp mp cos
p m¼1 m2
2 a2 a

The best fit is for DAB ¼ 9:0 10 6 cm2 =s, with an ESS ¼

ð18-51Þ
0.001669. The best fit of (18-49) is included as a line in Figure where for small values of DABt/z2, the infinite series term is
18.34. For values of N FoM > 0:1, only the first term in the infinite significant and converges very slowly.
series of (18-49) is significant, and therefore (18-49) approaches a
The average moisture content in the slab at any time dur-
straight line on a semilog plot, as can be observed for the theoretical
ing the constant-rate period is defined by
line in Figure 18.34, when t/a2 > 3.2 h/cm2.
Z
To determine the drying time for the 72-inch 12-inch 1-inch 1 a
poplar, assume that all drying takes place in the diffusion-controlled, X avg ¼ Xdz ð18-52Þ
a 0
falling-rate period, with mass transfer from the edges negligible and
a drying time long enough that N FoM > 0:1. Then (18-49) reduces to If (18-52) is integrated after substitution of X from (18-51),
X avg X 8 p 2
DAB t DAB rs DAB t
ln ¼ ln ð2Þ Xo X avg ¼ 2 ¼ N FoM ð18-53Þ
Xo X p2 4 a2 Rc a a
Solving (2) for (DABt/a2), From (18-51), it is seen that the generalized moisture
DAB t 4 8 Xo X profile during the constant-rate drying period, ðX o XÞ
N FoM ¼ ¼ 2 ln 2 ð3Þ DAB rs =ðRc aÞ, is a function of the dimensionless position
a2 p p X avg X
ratio, z/a, and N FoM , where the latter is equal to the general-
Xavg ¼ 0.10, Xo ¼ 0.45, and X ¼ 0.06 ized, average moisture content given by (18-53). A plot of
From (3), (18-51) for six position ratios, is given in Figure 18.35a.
" # Equation (18-51) is based on the assumption that during
4 8 0:45 0:06
N FoM ¼ ln ¼ 0:839 the constant-rate drying period, moisture will be supplied to
ð3:14Þ2 ð3:14Þ2 0:10 0:06
the surface by liquid diffusion at a rate sufficient to maintain
a constant moisture-evaporation rate. As discussed above, the
760 Chapter 18 Drying of Solids

1.0

Constant-rate drying
with evaporation at surface

=0
z/a
1–
0.1 ce,
r fa

_
= _1
u

8
1 , s
rve

z/a
(Xo – X) DAB ρs/Rca

= _1_
Cu

4
–
1

z/a

= _1_
2
1–

=1
z/a

_
z/a = _3
4
1–

– z/a
1–

ne, 1
0.01

la
Midp
0.001
0.0001 0.001 0.01 0.1 1.0
(NFoM) = DAB/ta2
(a) Moisture profile change

1.0
Constant-rate drying with
evaporation at surface
(Xo – Xavg)/(Xo – Xs)

0.1

0.01
0.01 0.1 1.0 10.0
(Xo – Xs)DAB ρs/Rca
(b) Surface moisture change

Figure 18.35 Changes in moisture concentration during constant-rate period while diffusion in the solid occurs.
[From W.H. Walker, W.K. Lewis, W.H. McAdams, and E.R. Gilliland, Principles of Chemical Engineering, 3rd ed., McGraw-Hill, New York (1937) with
permission.]

average moisture content at which the constant-rate period for the prediction is the assumption that the falling-rate
ends and the falling-rate period begins is called the critical period will begin when the moisture content at the surface
moisture content, Xc. In the empirical approach to the reaches the equilibrium-moisture content corresponding to
falling-rate period, Xc must be known from experiment for the conditions of the surrounding gas. This prediction is
the particular conditions because Xc is not a constant for a facilitated, as described by Walker et al. [21], by replotting
given material but depends on a number of factors, including an extension of Curve 1 in Figure 18.35a for the moisture
moisture diffusivity, slab thickness, initial- and equilibrium- content at the surface, Xs, in the form shown in Figure
moisture contents, and all factors that influence moisture 18.35b. Use of Figure 18.35b and the predicted influence of
evaporation in the constant-rate drying period. A useful several variables on the value of Xc is illustrated in the fol-
aspect of (18-51) is that it can be used to predict Xc. The basis lowing example.
 §18.4 Drying Periods 761

EXAMPLE 18.14 Critical Moisture Content. For a half-slab of thickness a, ms ¼ rs aA ð2Þ

Experiments by Gilliland and Sherwood [20] with brick clay mix rs a

Combining (1) and (2), tc ¼ ðX o Xc Þ ð3Þ
show that for certain drying conditions, moisture profiles conform Rc
reasonably well to the Case 2 diffusion theory. Use Figure 18.35b
For Set 1 of Example 18.14,
to predict the critical moisture content for the drying of clay slabs
from the two faces only under three different sets of conditions. For
ð1:6Þð0:5Þ
all three sets, Xo ¼ 0.30, X ¼ 0.05, rs ¼ 1:6 g=cm3 , and DAB ¼ tc ¼ ð0:30 0:11Þ ¼ 0:76 h
ð0:2Þ
0:3 cm2 =h. The other conditions are
DAB tc ð0:3Þð0:76Þ
Set 1 Set 2 Set 3 N FoM ¼ ¼ ¼ 0:91
a2 ð0:5Þ2
a, slab half-thickness, cm 0.5 0.5 1.0
Rc, drying rate in constant-rate 0.2 0.4 0.2 Because N FoM > 0:5, a parabolic profile is closely approached. Sim-
drying period, g/cm2-h ilarly, the following results are obtained for Sets 2 and 3:

Set 2 Set 3
Solution
tc, h 0.28 1.12
For Set 1, using Xs ¼ X ¼ 0.05, N FoM 0.34 0.34
DAB rs ð0:3Þð1:6Þ Parabolic profile no no
ðX o XsÞ ¼ ð0:30 0:05Þ ¼ 1:20 closely approached?
Rc a ð0:2Þð0:5Þ

X o X avg For Sets 2 and 3, the parabolic moisture-content profiles are not
From Figure 18.35b, ¼ 0:7
Xo Xs closely approached. However, the absolute errors in Xo X at the
surface and midplane are determined from (18-51) to be only 1.1%
Solving, X avg ¼ X c ¼ 0:25 0:7ð0:25 0:05Þ ¼ 0:11.
and 4.3%, respectively.
In a similar manner, Xc for Set 2 ¼ 0.16 and Xc for Set 3 ¼ 0.16.
These results show that doubling the rate of drying in the constant-
rate period or doubling the slab thickness substantially increases Xc. An approximate theoretical estimate of the additional dry-
ing time required for the falling-rate period is derived as fol-
For sufficiently large values of time, corresponding to lows from the development by Walker et al. [21]. At the end
N FoM ¼ DaAB2 t > 0:5, the term for the infinite series in (18-51) of the constant-rate period, the rate of flow of moisture by
approaches a value of 0, and, at all locations in the slab, X Fickian diffusion to the surface of the slab, where it is then
becomes a parabolic function of z. evaporated, may be equated to the reduction in average mois-
A simple equation for the parabolic distribution can be ture content of the slab. Thus,
formulated as follows from (18-51) in terms of the moisture rs aA dXavg dX
R¼ ¼ DAB rs s ð18-57Þ
contents at the surface and midplane of the slab. At the sur- A dt dz
face z ¼ a, the long-time form is From the parabolic moisture profile of (18-56), at the surface
z ¼ þ a,
Rc a 1
Xo Xs ¼ þ N FoM ð18-54Þ dX 2
DAB rs 3 ¼ ðX m XsÞ ð18-58Þ
Similarly, at the midplane, z ¼ 0, where X ¼ Xm, dz z¼þa a

Rc a 1 However, it is more desirable to convert this expression from

Xo Xm ¼ þ N FoM ð18-55Þ one in terms of Xm to one in terms of Xavg. To do this, (18-56)
DAB rs 6
can be substituted into (18-52) for the definition of Xavg, fol-
Combining (18-51), (18-54), and (18-55), the dimensionless lowed by integration to give
moisture-content profile becomes 2 1
X avg ¼ X m þ X s ð18-59Þ
Xm X z 2 3 3
¼ ð18-56Þ which can be rewritten as
Xm Xs a
3
Xm Xs ¼ X avg X s ð18-60Þ
2
EXAMPLE 18.15 Parabolic Moisture-Profile.
Substitution of (18-60) into (18-58), followed by substitution
For Example 18.14, determine the drying time for the constant-rate of the result into (18-57), gives
drying period and whether the parabolic moisture-content profile is
closely approached by the end of that period.
dX avg 3DAB rs
R¼ ars ¼ X avg Xs ð18-61Þ
dt a
Solution
The falling-rate period is assumed to begin with Xs ¼ X . If
From (18-39), ms ðX o X c Þ the parabolic moisture profile exists during the falling-rate
tc ¼ ð1Þ period and if Xs ¼ X remains constant, then (18-61) applies
ARc
during that period and the straight-line, falling-rate period
762 Chapter 18 Drying of Solids

shown in Figure 18.31a is obtained. Integrating (18-61) from Calculations for other values of time give the following results:
the start of the falling-rate period when Xavg ¼ Xc,
tf, Time from Start of Experimental Predicted
a2 Xc X Falling-Rate Period, minutes Xavg Xavg
tf ¼ ln ð18-62Þ
3DAB X avg X
0 0.165 0.165
Thus, the falling-rate-period duration is predicted to be 20 0.145 0.145
directly proportional to the square of the slab half-thickness 35 0.134 0.132
and inversely proportional to the moisture liquid diffusivity. 52 0.124 0.119
Equation (18-62) gives reasonable predictions for nonporous 71 0.114 0.106
slabs of materials such as wood, clay, and soap when the 95 0.106 0.093
slabs are thick and DAB is low. However, serious deviations 116 0.099 0.083
can occur when DAB depends strongly on X and/or tempera- 138 0.095 0.075
ture. In that case, an average DAB can be used to obtain an 149 0.090 0.071
approximate result. A summary of experimental average
moisture liquid diffusivities for a wide range of water-wet Comparing predicted values of Xavg with experimental values, the
solids is tabulated in Chapter 4 of Mujumdar [1]. deviation increases with increasing time. If the value of DAB is
reduced to 0.53 10 4 cm2/s, much better agreement is obtained
with the ESS decreasing from 0.0013 to 0.000154 cm4/s2.
EXAMPLE 18.16 Falling-Rate Period in Drying.
Gilliland and Sherwood [20] obtained data of the drying of water-
wet 7 7 2:54-cm slabs of 193.9 g (bone-dry) brick clay mix for Capillary-Flow Theory
direct-heat convective air drying from the two faces in both the
constant- and falling-rate periods. For Xo ¼ 0.273, X ¼ 0.03, the For wet solids of the first category, as discussed in §18.3,
rate of drying in the constant-rate period to Xc ¼ 0.165 was 0.157 g/ moisture is held as free moisture in the interstices of the par-
h-cm2. The air velocity past the two faces was 15.2 m/s, with ticles. Movement of moisture from the interior to the surface
Td ¼ 25 C and Tw ¼ 17 C. During the falling-rate period, experi- can occur by capillary action in the interstices, but may be
mental average slab moisture contents were as follows: opposed by gravity.
Cohesive forces hold liquid molecules together. Also,
liquid molecules may be attracted to a solid surface by adhe-
Time from Start of the
Constant-Drying sive forces. Thus, water in a glass tube will creep up the side
Rate Period, minutes X avg of the tube until adhesive forces are balanced by the weight
of the liquid. For an ideal case of a capillary tube of small
67 0.165 (critical value) diameter partially immersed vertically in a liquid, the liquid
87 0.145 rises in the tube to a height above the surface of the liquid in
102 0.134 the reservoir. At equilibrium, the height, h, will be
119 0.124
138 0.114 h ¼ 2s=rL gr ð18-63Þ
162 0.106
where s is the surface tension of the liquid and r is the radius
183 0.099
of the capillary. The smaller the radius of the capillary, the
205 0.095
larger the capillary effect. Unlike mass transfer by diffusion,
216 0.090
which causes moisture to move from a region of high to low
concentration, liquid in interstices flows because of capillary
At the end of the constant-drying-rate period, the moisture profile effects, regardless of concentration.
is assumed parabolic. Other experiments give DAB ¼ 0:72 For capillary flow in granular beds of wet solids, the varia-
10 4 cm2 =s. ble size and shape of the particles make it extremely difficult
Use (18-62) to predict values of Xavg during the falling-rate pe- to develop a usable theory for predicting the rate of drying in
riod and compare predicted values to experimental values. the falling-rate period in terms of permeability and capillar-
ity. Interesting discussions and idealized theories are pre-
Solution
sented by Keey [7, 23] and Ceaglske and Kiesling [22], but
Solving (18-62), for practical calculations, it appears that, despite pleas to the
X avg ¼ X þ ðX c X Þexpð 3DAB tf =a2 Þ ð1Þ contrary, it is common to apply diffusion theory with effec-
tive diffusivities determined from experiment. In general,
where tf is the time from the start of the falling-rate period. these diffusivities are lower than those for true diffusion of
For tf ¼ 87 67 ¼ 20 min, from (1),
moisture in nonporous materials. Some values are included
X avg ¼ 0:03 þ ð0:165 0:03Þ in a tabulation in Chapter 4 of Mujumdar [1]. For example,
exp½ 3ð0:72 10 4 Þð20Þð60Þ=ð1:27Þ2 ¼ 0:145 cm2 =s effective diffusivities of water in beds of sand particles cover
a range of 1.0 10 2 to 8:0 10 4 cm2 =s.
 §18.5 Dryer Models 763

§18.5 DRYER MODELS where any convenient reference temperatures can be used to
determine the enthalpies.
Previous sections developed general mathematical models
When the system is air, water, and a solid, a more conve-
for estimating drying rates and moisture profiles for batch
nient form of (18-67) can be obtained by evaluating the
tray dryers of the cross-circulation and through-circulation
enthalpies of the solid and the air from specific heats, and
types. More specific models for continuous dryers have been
obtaining moisture enthalpies from the steam tables. Often,
developed over the years, and this section presents three of
the specific heat of the solid is almost constant over the tem-
them: (1) belt dryer with through-circulation, (2) direct-heat
perature range of interest, and in the range from 25 C (78 F)
rotary dryer, and (3) fluidized-bed dryer, all of which are cat-
to 400 C (752 F), the specific heat of dry air increases by less
egorized as direct-heat dryers. Other models are considered
than 3%, so the use of a constant value of 0.242 Btu/lb- F
by Mujumdar [1] and in a special issue of Drying Technol-
introduces little error. If the enthalpy reference temperature of
ogy, edited by Genskow [24].
the water is taken as To (usually 0 C (32 F) for liquid water
when using the steam tables), (18-67) can be rewritten as
§18.5.1 Material and Energy Balances
ms ðCP Þs ðT ds T ws Þ þ X ds ðH H2 O Þds X ws ðH H2 O Þws
for Direct-Heat Dryers
¼ mg ½ðC PÞair T gi T go þ gi ðH H2 O Þgi ð18-68Þ
Consider the continuous, steady-state, direct-heat dryer
shown in Figure 18.36. Although countercurrent flow is go ðH H2 O Þgo

shown, the following development applies equally well to A further simplification in the energy balance for the air–
cocurrent flow and crossflow. Assume that the dryer is per- water–solid system can be made by replacing enthalpies for
fectly insulated so that the operation is adiabatic. As the solid water by their equivalents in terms of specific heats for liquid
is dried, moisture is transferred to the gas. No solid is water and steam and the heat of vaporization. In the range
entrained in the gas, and changes in kinetic energy and poten- from 25 C (78 F) to 100 C (212 F), the specific heat of
tial energy are negligible. The flow rates of dry solid, ms, and liquid water and steam are almost constant at 1 Btu/lb- F and
dry gas, mg, do not change as drying proceeds. Therefore, a 0.447 Btu/lb- F, respectively. The heat of vaporization of
material balance on the moisture is water over this same range decreases from 1,049.8 to
X ws ms þ gi mg ¼ X ds ms þ go mg ð18-64Þ 970.3 Btu/lb, a change of almost 8%. Combining (18-65)
with (18-68) and taking a thermodynamic path of water evap-
The rate of moisture evaporation, my, is given by a re-
oration at the moisture-evaporation temperature, denoted Ty,
arrangement of (18-64):
the simplified energy balance is
my ¼ ms ðX ws X ds Þ ¼ mg H go H gi ð18-65Þ n
where the subscripts are ws (wet solid), ds (dry solid), gi (gas Q ¼ ms ðC P Þs ðT ds T ws Þ þ X ws ðC P ÞH2 Oð‘Þ ðT y T ws Þ
in), and go (gas out). þX ds ðCP ÞH2 Oð‘Þ ðT ds T y Þ
An energy balance can be written in terms of enthalpies or h io
in terms of specific heat and heat of vaporization. In either þðX ws Xds Þ DH vap y þ ð C P Þ H2 OðgÞ T go T y
case, it is convenient to treat the dry gas, dry solid, and mois- nh i o
ture (liquid and vapor) separately, and assume ideal mixtures. ¼ mg ðC PÞair þ gi ðC P ÞH2 OðyÞ T gi T go
In terms of enthalpies, the energy balance is as follows,
ð18-69Þ
where s, g, and m refer, respectively, to dry solid, dry gas,
and moisture: Equations (18-64)–(18-69) are useful for determining the
required gas flow rate for drying a given flow rate of wet
ms ðH s Þws þX ws ms ðH m Þws þ mg H g gi
þ gi mg ðH m Þgi solids, as illustrated in the next example. Also of interest for
¼ ms ðH s Þds þ X ds ms ðH m Þds ð18-66Þ sizing the dryer is the required heat-transfer rate, Q, from the
gas to the solid. For the air–water–solid system, this rate is
þ mg H g go
þ go mg ðH m Þgo
equal to either the LHS or the RHS of (18-69), as indicated.
A factored rearrangement of (18-66) is In the general case,
n h io
ms ðH s Þds ðH s Þws þ X ds ðH m Þds X ws ðH m Þws Q ¼ mg H g gi H g go þ gi ðH m Þgi ðH m Þgo
h
¼ mg H g gi H g go þ gi ðH m Þgi ð18-67Þ ¼ ms ðH s Þds ðH s Þws þ X ds ðH m Þds X ws ðH m Þws
i
go ðH m Þgo
þ mg ½ðH m Þgo go gi ð18-70Þ

Exiting gas, go Entering gas, gi

mg go
Adiabatic, mg gi
continuous,
direct-heat
dryer
Wet solid feed, ws Dry solid product, ds Figure 18.36 General configuration for a
ms Xws ms Xds continuous, direct-heat dryer.
764 Chapter 18 Drying of Solids

EXAMPLE 18.17 Balance for a Direct-Heat Dryer.
A continuous, cocurrent-flow direct-heat dryer is to be used to
dry crystals of Epsom salt (magnesium sulfate heptahydrate).
The feed to the dryer, a filter cake from a rotary, vacuum filter,
is 2,854 lb/h of crystals (dry basis) with a moisture content of
25.8 wt% (dry basis) at 85 F and 14.7 psia. Air enters at 14.7
psia, with dry-bulb and wet-bulb temperatures of 250 F and
117 F. The final moisture content of the dried crystals is to be
1.5 wt% (dry basis), at no more than 118 F to prevent decompo-
sition of the heptahydrate (see Figure 17.2). Determine: (a) rate
of moisture evaporation, (b) outlet temperature of the air, (c)
rate of heat transfer, and (d) entering air flow rate. The average
specific heat of Epsom salt is 0.361 Btu/lb- F.
Solution
ms ¼ 2;854 lb=h; X ws ¼ 0:258; X ds ¼ 0:015; T ws ¼ 85 F;
T ds ¼ 118 F; T gi ¼ 250 F; and T y ¼ 117 F
From Figure 18.17 for Tdb ¼ 250 F and Twb ¼ 117 F, gi ¼ 0:0405.
(a) From (18-65), my ¼ 2;854ð0:258 0:015Þ ¼ 694 lb=h.
(b) Because the dryer operates cocurrently, the outlet temperature of
the gas must be greater than the outlet temperature of the dry
solid, which is taken as 118 F. The best value for Tgo is obtained
by optimizing the cost of the drying operation. A reasonable
value for Tgo can be estimated by using the concept of the num-
ber of heat-transfer units, which is analogous to the number of
transfer units for mass transfer, as developed in §6.7. For heat
transfer in a dryer, where the solids temperature throughout most
of the dryer will be at Ty, the number of heat-transfer units is
§18.5.2 Belt Dryer with Through-Circulation
Consider the continuous, two-zone through-circulation belt
dryer in Figure 18.37a. A bed of wet-solid particles is con-
veyed continuously into Zone 1, where contact is made with
hot gas passing upward through the bed. Because the temper-
ature of the gas decreases as it passes through the bed, the
temperature-driving force decreases so that the moisture con-
tent of solids near the bottom of the moving bed decreases
more rapidly than for solids near the top. To obtain a dried
solid of more uniform moisture content, the gas flow direc-
tion through the bed is reversed in Zone 2.
Based on the work of Thygeson and Grossmann [25], a
mathematical model for Zone 1 can be developed using the
coordinate system shown in Figure 18.37b, based on six
assumptions:
1. Wet solids enter Zone 1 with a uniform moisture con-
tent of Xo on the dry basis.
2. Gas passes up through the moving bed in plug flow
with no mass transfer in the vertical direction (i.e., no
axial dispersion).
Exit gas 1 Hot gas in 2
Wet solids Dry solids
Zone 1 Zone 2

T gi Ty
N T ¼ ln ð1Þ
T go Ty

where economical values of NT are usually in the range of 1.0–

2.5. Assume a value of 2.0. From (1),
Hot gas in 1 Exit gas 2
250 117 (a) Configuration
2 ¼ ln
T go 117
from which Tgo ¼ 135 F. Gas out

(c) The rate of heat transfer is obtained from (18-69) using the con-
ditions for the solid flow.
z=H
Q ¼ 2;854f0:361ð118 85Þ þ 0:258ð1 Þð117 85Þ
þ 0:015ð1Þð118 117Þ þ ð0:258 0:015Þ
½1;027:5 þ ð0:447Þð135 117Þ g
¼ 2;854½11:9 þ 8:3 þ 0:02 þ 249:7 þ 2:0 Partially dried
¼ 2;854ð271:9Þ ¼ 776;000 Btu=h Wet solids solids
dz
Solids
dx
It should be noted that the heat required to vaporize the 694 lb/h
of moisture at 117 F is (249.7/271.9) 100% ¼ 91.8% of the Gas
total heat load.
z
(d) The entering air flow rate is obtained from (18-69) using the far
RHS of that equation with the above value of Q. z=0
x
776;000 x=0 x = L1
mg ¼ ¼ 25;940 lb=h
½ð0:242Þ þ ð0:0405Þð0:447Þ ð250 135Þ
Gas in
The total entering air, including the humidity, is 25,940 (b) Coordinate system for zone 1
(1 þ 0.0405) ¼ 27,000 lb/h.
Figure 18.37 Continuous, two-zone through-circulation belt dryer.
 §18.5 Dryer Models 765

3. Drying takes place in the constant-rate period, con- Zone 2

trolled by the rate of heat transfer from the gas to the
In Zone 2, (18-71) still applies, with X1 and T1 replaced by X2
surfaces of the solid particles, where the temperature is
and T2, but the initial condition for Xo is ðX 1 ÞL1 from (18-75)
the adiabatic-saturation temperature.
for x ¼ L1, which depends on z. Equation (18-72) also
4. Sensible-heat effects are negligible compared to latent- applies, with T1 replaced by T2, but the initial condition is
heat effects. T2 ¼ Tgi at z ¼ H. The integrated result is
5. The void fraction of the bed is uniform and constant, " #
and no mixing of solid particles occurs. h aðH zÞ
T 2 ¼ T y þ T gi T y exp ð18-77Þ
rg ðC P Þg us
6. The solids are conveyed at a uniform linear speed, S.
With T go2 given by (18-74), where T go1 is replaced by T go2 ,
Based on these assumptions, the gas temperature
and
decreases with increasing distance z from the bottom of the " #
bed, and is independent of the distance, x, in the direction in h aðH zÞ
which the solids are conveyed, i.e., T ¼ T{z}. The moisture xh a T gi T y exp
X2 rg ðCP Þg us
content of the solids varies in both z- and x-directions, ¼1 vap ð18-78Þ
decreasing more rapidly near the bottom of the bed, where ðX 1 ÞL1 SDH y ðrb Þds
the gas temperature is higher, i.e., X ¼ X{z, x}. where ðX 1 ÞL1 is the value from (18-75) for x ¼ L1 at the value
of z in (18-78). The value of x in (18-78) is the distance from
Zone 1 the start of Zone 2. Values of ðX 2 ÞL2 at any z are obtained
from (18-76) for x ¼ L2. The average moisture content over
With no mixing of solids, a material balance on the moisture the height of the moving bed leaving Zone 2 is then obtained
in the solids at any vertical location, z, is given by from (18-76), with ðX 1 ÞL1 replaced by ðX 2 ÞL2 . The above rela-
dX1 dX 1 ðhaÞðT 1 T y Þ tionships are illustrated in the next example.
¼S ¼ ð18-71Þ
dt dx DH vap
y ðrb Þds
where a ¼ surface area of solid particles per unit volume of EXAMPLE 18.18 Through-Circulation Drying.
bed; T1 ¼ bulk temperature of the gas in Zone 1, which
depends on z; and (rb)ds is the bulk density of solids when The filter cake of CaCO 3 in Example 18.8 is to be dried continu-
dry. The initial condition is X1 ¼ Xo for x ¼ 0. ously on a belt dryer using through-circulation. The dryer is 6 ft
An energy balance for the gas phase at any location x is wide, has a belt speed of 1 ft/minute and consists of two drying
zones, each 12 ft long. Air at 170 F and 10% relative humidity
dT 1 enters both zones, passing upward through the bed in the first zone,
rg ðCP Þg us ¼ ðhaÞðT 1 T y Þ ð18-72Þ
dz and downward in the second, at a superficial velocity of 2 m/s. Bed
where rg ¼ gas density and us ¼ superficial velocity of gas height on the belt is 2 inches. Predict the moisture-content distribu-
through the bed. The initial condition is T1 ¼ Tgi for z ¼ 0. tion with height at the end of each zone, and the average moisture
Equation (18-71) is coupled to (18-72), which is indepen- content at the end of Zone 2. Assume all drying is in the constant-
rate period and neglect preheat.
dent of (18-71). It is possible to separate variables and inte-
grate (18-72) to obtain
! Solution
h az
T 1 ¼ T y þ T gi T y exp ð18-73Þ From data in Examples 18.7 and 18.8,
rg ðC P Þg us
X o ¼ 0:30
At z ¼ H at the top of the bed, 1:00
! ðrb Þds ¼ ð1;950Þ ¼ 1;500 kg=m3
1:30
h aH eb ¼ 0:50
T go1 ¼ T y þ T gi T y exp ð18-74Þ
rg ð C P Þg us
T y ¼ 37:8 C ¼ 311 K; T gi ¼ 76:7 C ¼ 350 K
Equation (18-71) can now be solved by combining it with DH vap
y ¼ 2;413 kJ=kg
(18-73) to eliminate T1, followed by separation of variables
From extrusion area and volume in Example 18.8,
and integration. The result is
! 3:16 10 4 ð0:5Þ
h az a¼ ¼ 395 m2 =m3 bed
x ha T gi T y exp 4:01 10 7
X1 rg ðCP Þg us
¼1 vap ð18-75Þ For us ¼ 2 m/s, h ¼ 0.188(kJ/s-m2-K2) from Example 18.8.
Xo SDH y ðrb Þds
ðC P Þg ¼ 1:09 kJ=kg-K; rg at 1 atm ¼ 0:942 kg=m3 ;
The moisture content ðX 1 ÞL1 at x ¼ L1 is obtained by replacing S ¼ 1 ft=min ¼ 0:00508 m=s
x with L1. If desired, Xavg at x ¼ L1 can be determined from
Z H The cross-sectional area of the moving bed normal to the conveying
direction is 6(2/12) ¼ 1 ft2 ¼ 0.0929 m2. For a belt speed of 1 ft/min¼
X avg ¼ ðH 1 ÞL1 dz ð18-76Þ
0
0.305 m/minute, the volumetric flow of solids is (0.0929)(0.305)
766 Chapter 18 Drying of Solids

¼ 0.0283 m3/minute. The mass rate of flow is 0.0283(1,500) ¼ A commercial-size direct-heat rotary dryer should be
42.5 kg/min (dry basis). scaled up from pilot-plant data. However, if a representative
Zone 1 sample of the wet solid is not available, the following proce-
dure and model, based on test results with several materials
H ¼ 0:167 ft ¼ 0:0508 m and L1 ¼ 12 ft ¼ 3:66 m
in both pilot-plant-size and commercial-size dryers, is useful
From (18-74), the gas temperature leaving the bed is for a preliminary design.
The hot gas can flow countercurrently or cocurrently to
ð0:188Þð395Þð0:0508Þ
T go1 ¼ 37:8 þ ð76:7 37:8Þexp the flow of the solids. Cocurrent flow is used for very wet,
ð0:942Þð1:09Þð2Þ
sticky solids with high inlet-gas temperatures, and for non-
¼ 44 C ¼ 317 K hygroscopic solids. Countercurrent flow is preferred for low-
The moisture-content distribution at x ¼ L1 is obtained from to-moderate inlet-gas temperatures, where thermal efficiency
(18-75). For z ¼ 0, becomes a factor. When solids are not subject to thermal deg-
ð3:66Þð0:188Þð395Þð76:7 37:8Þ radation, melting, or sublimation, an inlet-gas temperature up
X 1 ¼ 0:30 1 ¼ 0:127 to 1,000 F can be used. The exit-gas temperature is deter-
ð0:00508Þð2;413Þð1;500Þ
mined from economics, as discussed in Example 18.17,
For other values of z, a spreadsheet gives: where Equation (1) can be used with NT in the range of 1.5–
2.5. Generally, more gas flow and higher gas temperatures in-
z, m ðX 1 ÞL1 crease operating costs, but decrease capital costs, because the
0 0.127 larger temperature-driving force increases the heat-transfer
0.0127 0.191 rate. Allowable gas velocities are determined from the dust-
0.0254 0.231 ing characteristics of the particles, and can vary widely with
0.0381 0.257 particle-size distribution and particle density. Some typical
0.0508 0.273 values for allowable gas velocity are as follows:

Average Allowable
Because of the decrease in gas temperature as it passes through the
Particle Particle Gas
bed, moisture content varies considerably over the bed depth.
Density, Size, Velocity,
Zone 2 Material rp, lb/ft3 dp, mm uall, ft/s
The flow of air is reversed to further the drying and smooth out the
moisture-content distribution. The value of Xo is replaced by the Plastic granules 69 920 3.5
above values of ðX 1 ÞL1 for corresponding values of z. Using (18-78) Ammonium nitrate 104 900 4.5
with a spreadsheet, the following distribution is obtained at Sand 164 110 1.0
L2 ¼ 3.66 m for a total length of both zones ¼ 24 ft ¼ 7.32 m:
Sand 164 215 2.0
Sand 164 510 5.0
z, m ðX 2 ÞL 2
Sawdust 27.5 640 1.0
0 0.116
0.0127 0.163 Using an appropriate allowable gas velocity, uall, with mass
0.0254 0.178 flow rate and density of the gas at the gas-discharge end,
0.0381 0.163 (mg)exit, and (rg)exit, the dryer diameter, D, can be estimated
0.0508 0.116 by the continuity equation
" #0:5
4 mg exit
A much more uniform moisture distribution is achieved. From D¼ ð18-79Þ
puall rg exit
(18-76) for Zone 2, using numerical integration with a spreadsheet,
(X2)avg ¼ 0.155. Residence time of the solids in the dryer, u, is related to the
fractional volume holdup of solids, VH, by
LV H
u¼ ð18-80Þ
§18.5.3 Direct-Heat Rotary Dryer FV
where L ¼ length of dryer cylinder and FV ¼ solids volumetric
As discussed by Kelly in Mujumdar [1], design of a direct-
velocity in volume/unit cross-sectional area-unit time. A con-
heat rotary dryer, of the type shown in Figure 18.7, for drying
servative estimate of the holdup, including the effect of gas
solid particles at a specified feed rate, initial moisture content
velocity, is obtained by combining (18-80) with a relation in [2]:
Xws, and final moisture content Xds, involves determination of
heating-gas inlet and outlet conditions, heating-gas velocity 0:23F V G 5=d 0:5
p
and flow direction, dryer-cylinder diameter and length, VH ¼ 0:6 ð18-81Þ
SN 0:9 D rp
dryer-cylinder slope and rotation rate, number and type of
lifting flights, solids holdup as a % of dryer-cylinder volume, where FV ¼ ft3 solids/(ft2 cross section)-h; S ¼ dryer-cylinder
and solids-residence time. slope, ft/ft; N ¼ dryer-cylinder rate of rotation, rpm; D ¼
 §18.5 Dryer Models 767

dryer diameter, ft; G ¼ gas superficial mass velocity, lb/h-ft2; Solution

and dp ¼ mass-average particle size, mm. The plus (þ) sign
From the psychrometric chart (Figure 18.17), Twb ¼ 107 F. Assume
on the second term corresponds to countercurrent flow that
that all drying is at this temperature for the solid. A reasonable out-
tends to increase the holdup, while the minus ( ) sign let temperature for the air can be estimated from (1) of Example
denotes cocurrent flow. Equation (18-81) holds for dryers hav- 18.17, assuming NT ¼ 1.5. From that equation,
ing lifting flights with lips, but is limited to gas velocities less
250 107
than 3.5 ft/s. A more complex model by Matchett and Sheikh 1:5 ¼ ln
T go 107
[26] is valid for gas velocities up to 10 ft/s. Optimal solids
holdup is 10–18% of dryer volume so that flights run full and Solving, Tgo ¼ 140 F. Assume solids outlet temperature ¼ Tds ¼
all or most of the solids are showered during each revolution. 135 F.
When drying is in the constant-rate period such that the Heat-transfer rate:
rate can be determined from the rate of heat transfer from the
ms ¼ 700ð60Þ ¼ 42;000 lb=h of solids ðdry basisÞ;
gas to the wet surface of the solids at the wet-bulb tempera-
ture, a volumetric heat-transfer coefficient, ha, can be used, ðCP Þs ¼ 0:4 Btu=lb- F; T ws ¼ 70 F; T y ¼ T wb ¼ 107 F;
which is defined by X ws ¼ 0:15; X ds ¼ 0:01; and DH vap
y ¼ 1;033 Btu=lb
Q ¼ ðhaÞVDT LM ð18-82Þ
From (18-65), my ¼ 42;000ð0:15 0:01Þ ¼ 5;880 lb=h H2 O evaporated.
where V ¼ volume of dryer cylinder ¼ pD2L/4; From (18-69),
Tg in
Tg out#
Q ¼ 42;000fð0:4Þð135 70Þ þ ð0:15Þð1Þð107 70Þ
DT LM ¼ " ð18-83Þ
Tg Ty þð0:01Þð1Þð135 70Þ þ ð0:15 0:01Þ½1;033 þ ð0:447Þð140 107Þ g
in
ln ¼ 7;510;000 Btu=h
Tg out
Ty
Air flow rate:
and ha ¼ volumetric heat-transfer coefficient based on dryer-
7;510;000
cylinder volume as given by the empirical correlation of mg ¼
½ð0:242Þ þ ð0:02Þð0:447Þ ð250 135Þ
McCormick [27], when the heating gas is air:
¼ 260;000 lb=h entering dry air
ha ¼ KG0:67 =D ð18-84Þ
Dryer diameter:
where ha is in Btu/h-ft 3- F, G is in lb/h-ft2, and D is in ft.
Assume an allowable gas velocity at the dryer exit of 4.5 ft/s.
K ¼ 0.5 is recommended in [2] for dryers operating at a
peripheral cylinder speed of 1.0–1.25 ft/s and with a flight ðmg Þexit ¼ 260;000ð1 þ 0:02Þ þ 5;880 ¼ 271;000 lb=h total gas
count of 2.4D to 3.0D per circle. When K is available from PM
rg ¼
pilot-plant data, (18-84) can be used for scale-up to a larger exit RT go
diameter and a different value of G.
It might be expected that a correlation for the volumetric 271;000
M ¼ ¼ 28:3
260;000 11;000
heat-transfer coefficient, ha, would take into account the parti- þ
29 18
cle diameter because the solids are lifted and showered
ð1Þð28:3Þ
through the gas. However, the solids shower as curtains of rg ¼ ¼ 0:0646 lb=ft3
exit ð0:730Þð600Þ
some thickness, with the gas passing between the curtains.
Thus, particles inside the curtains do not receive significant 4ð271;000Þ 0:5
exposure to the gas, and the effective heat-transfer area is From (18-79), D¼ ¼ 18 ft
ð3:14Þð4:5Þð3;600Þð0:0646Þ
more likely determined by the areas of the curtains. Neverthe-
less, (18-84) accounts for only two of the many possible varia- Dryer length:
bles, and the inverse relation with dryer diameter is not well ð271;000Þð4Þ
G ¼ Gexit ¼ ¼ 1;070 lb=h-ft2
supported by experimental data. A complex model for heat ð3:14Þð18Þ2
transfer that treats h and a separately is that of Schofield and
From (18-84), ha ¼ 0:5ð1;070Þ0:67 =18 ¼ 3 Btu=h-ft3 - F.
Glikin [28], as modified by Langrish, Bahu, and Reay [29].
From (18-83), neglecting the periods of wet solids heating up to
107 F and the dry solids heating up to 135 F, because the heat trans-
EXAMPLE 18.19 Direct-Heat Rotary Dryer. ferred is a small % of the total,

Ammonium nitrate, at 70 F with a moisture content of 15 wt% (dry 250 140

basis), is fed into a direct-heat rotary dryer at a feed rate of 700 lb/ DT LM ¼ ¼ 75 F
250 107
minute (dry basis). Air at 250 F and 1 atm, with a humidity of 0.02 ln
140 107
lb H2O/lb dry air, enters the dryer and passes cocurrently with the
solid. The final solid moisture content is to be 1 wt% (dry basis) and 7;510;000
all drying will take place in the constant-rate period. Make a prelim- From (18-82), V¼ ¼ 33;400 ft3
ð3Þð75Þ
inary estimate of the dryer diameter and length, assuming that such
Cross-sectional area ¼ (3.14)(18)2/4 ¼ 254 ft2.
dryers are available in: (1) diameters from 1 to 5 ft in increments of
0.5 ft and from 5 to 20 ft in increments of 1.0 ft, and (2) lengths 33;400
L¼ ¼ 130 ft
from 5 to 150 ft in increments of 5 ft. 254
768 Chapter 18 Drying of Solids

§18.5.4 Fluidized-Bed Dryer ! ! !

DP Cross-sectional Volume
The behavior of a bed of solid particles when a gas is passed across area ¼ of
bed of bed bed
up through the bed is shown in Figure 18.38. At a very low 0 120 1 0 13
gas velocity, the bed remains fixed. At a high gas velocity, Volume Density Density
the bed disappears; the particles are pneumatically trans- B fraction C6B of C B of C7
@ of solid A4@ solid A @ displaced A5
ported by the gas when its local velocity exceeds the particle
particles particles gas
terminal settling velocity. and the system becomes a ‘‘gas
lift.’’ At an intermediate gas velocity, the bed is expanded,
but particles are not carried out by the gas. Such a bed is Thus, DPb Ab ¼ Ab Lb ð1 eb Þ½ðrp rg Þg ð18-86Þ
said to be fluidized, because the bed of solids takes on the The minimum gas-fluidization superficial velocity, umf, is
properties of a fluid. Fluidization is initiated when the gas obtained by solving (18-85) and (18-86) simultaneously for
velocity reaches the point where all the particles are sus- u ¼ umf. For N Re;p ¼ d p umf rg =m < 20, the turbulent-flow
pended by the gas. As the gas velocity is increased further, contribution to (18-85) is negligible and the result is
the bed expands and bubbles of gas are observed to pass up
through the bed. This regime of fluidization is referred to as d 2p rp rg g e3b f2s
bubbling fluidization and is the most desirable regime for umf ¼ ð18-87Þ
most fluidized-bed operations, including drying. If the gas 150m 1 eb
velocity is increased further, a transition to slugging fluidiza- For operation in the bubbling fluidization regime, a
tion eventually occurs, where bubbles coalesce and spread to superficial-gas velocity of us ¼ 2umf is a reasonable choice.
a size that approximates the diameter of the vessel. To some At this velocity, the bed will be expanded by about 10%,
extent, this behavior can be modified by placing baffles and with no further increase in pressure drop across the bed. In
low-speed agitators in the bed. this regime, the solid particles are well mixed and the bed
Before fluidization occurs, when the bed of solids is temperature is so uniform that fluidized beds are used indus-
fixed, the pressure drop across the bed for gas flow, DPb, trially to calibrate thermocouples and thermometers. If the
is predicted by the Ergun [30] equation, discussed in fluidized bed is operated continuously at steady-state condi-
§6.8.2: tions rather than batchwise with respect to the particles, the
particles will have a residence-time distribution like that of a
DPb ð1 eb Þ2 mus ð1 eb Þ rg u2s fluid in a continuous-stirred-tank reactor (CSTR). Some par-
¼ 150 2 þ 1:75
Lb e3b fs d p e3b fs d p ticles will be in the dryer for only a very short period of time
ð18-85Þ and will experience almost no decrease in moisture content.
Other particles will be in the dryer for a long time and may
where Lb ¼ bed height, us ¼ superficial-gas velocity, and fs ¼ come to equilibrium before that time has elapsed. Thus, the
particle sphericity. The first term on the RHS is dominant at dried solids will have a distribution of moisture content. This
low-particle Reynolds numbers where streamline flow exists, is in contrast to a batch-fluidization process, where all parti-
and the second term dominates at high-particle Reynolds cles have the same residence time and, therefore, a uniform
numbers where turbulent flow exists. final moisture content. This is an important distinction
The onset of fluidization occurs when the drag force on because continuous, fluidized-bed dryers are usually scaled
the particles by the upward-flowing gas becomes equal to the up from data obtained in small, batch fluidized-bed dryers.
weight of the particles (accounting for displaced gas): Therefore, it is important to have a relationship between
batch drying time and continuous drying time.
The distribution of residence times for effluent from a per-
fectly mixed vessel operating at continuous, steady-state con-
Fixed bed Bubbling Slugging Transport
fluidized fluidized bed ditions is given by Fogler [31] as
bed bed
Fluid Eftg ¼ expð t=tÞ=t ð18-88Þ
plus
Fluid Fluid Fluid particles where t is the average residence time and E{t} is defined
such that E{t}dt ¼ the fraction ofR t effluent with a residence
time between t and t þ dt. Thus, 01 Eftgdt ¼ fraction of the
effluent with a residence time less than t1. For example, if the
average particle-residence time is 10 min, 63.2% of the parti-
cles will have a residence time of less than 10 minutes. If the
particles are small and nonporous such that all drying takes
place in the constant-rate period, and u is the time for com-
plete drying, then
Very low Intermediate Higher High X t
velocity velocity velocity velocity ¼1 ; t u ð18-89Þ
Xo u
Figure 18.38 Regimes of fluidization of a bed of particles by a gas.
 §18.5 Dryer Models 769

The average moisture content of the dried solids leaving From (18-65), my ¼ 8;330ð0:20 0:05Þ ¼ 1;250 lb=h evaporated
the fluidized-bed is obtained by integrating the expression moisture.
below from 0 to only u because X ¼ 0 for t > u.
ð7;170Þð0:01Þ þ 1;250
Z u Z u go ¼ ¼ 0:184 lb H2 O=lb dry air
t 7;170
X ds ¼ XEftgdt ¼ X0 1 Eftgdt ð18-90Þ
0 0 u Total exiting gas flow rate ¼ 7,170(1 þ 0.184) ¼ 8,500 lb/h
Combining (18-88) and (18-90) and integrating gives
Minimum fluidization velocity:
1 expð u=tÞ
X ds ¼ X o 1 ð18-91Þ 8;500
u=t M of existing gas ¼ ¼ 26:5
7;170 1;330
If the particles are porous and without surface moisture such þ
29 18
that all drying takes place in the falling-rate period, diffusion
PM ð1Þð26:5Þ
theory may apply such that the following empirical exponen- ðrg Þexit ¼ ¼ ¼ 0:060 lb=ft3
RT g ð0:730Þð605Þ
tial expression may be used for the moisture content as a
function of time: ¼ 0:00096 g=cm3
X
¼ expð BtÞ ð18-92Þ m ¼ 0:048 lb=ft-h ¼ 0:00020 g=cm-s
Xo
In this case, the combination of (18-92) with (18-90), fol- For small particles, assume streamline flow at umf so that (18-87)
lowed by integration from t ¼ 0 to t ¼ 1, gives applies, but check to see if N Re;p < 20. Using cgs units,

X ds ¼ 1=ð1 þ BtÞ ð18-93Þ ð0:0500Þ2 ð2:6 0:00096Þð980Þð0:55Þ3 ð0:67Þ2

umf ¼
Values of u and B are determined from experiments with lab- 150ð0:00020Þð1 0:55Þ
oratory batch fluidized-bed dryers for scale-up to large dryers ¼ 35:3 cm=s
operating under the same conditions. d p umf rg ð0:0500Þð35:3Þð0:00096Þ
N Re; p ¼ ¼ ¼ 8:5
m 0:00020

EXAMPLE 18.20 Fluidized-Bed Dryer. Since NRe,p < 20, (18-87) does apply.

Ten thousand lb/h of wet sand at 70 F with a moisture content of Use a superficial-gas velocity of twice umf ¼ 2ð35:3Þ ¼
20% (dry basis) is to be dried to a moisture content of 5% (dry basis) 70:6 cm=s ¼ 8;340 ft=h.
in a continuous, fluidized-bed dryer operating at a pressure of 1 atm Bed diameter:
in the free-board region above the bed. The sand has a narrow size
range, with an average particle size of 500 mm; a sphericity, fs, of Equation (18-79) applies:
0.67; and a particle density of 2.6 g/cm3. When the sand bed is dry, 0:5
4ð8;500Þ
its void fraction, eb, is 0.55. Fluidizing air will enter the bed at a D¼ ¼ 4:7 ft
temperature of 1,000 F with a humidity of 0.01 lb H2O/lb dry air. 3:14ð8;340Þð0:060Þ
The adiabatic-saturation temperature is estimated to be 145 F. Batch Bed density:
pilot-plant tests with a fluidization velocity of twice the minimum
show that drying takes place in the constant-rate period and that all Fixed-bed density ¼ rs(1 eb) ¼ 2.6(1 0.55)(62.4) ¼ 73.0 lb/ft 3.
moisture can be removed in 8 minutes using air at the same condi- Assume the bed expands by 10% upon fluidization to u ¼ 2umf:
tions and with a bed temperature of 145 F. Determine the bed height 73:0
and diameter for the large, continuous unit. rb ¼ ¼ 66 lb=ft3 ðdry basisÞ
1:10
Average particle-residence time:
Solution
d p ¼ 500 mm ¼ 0:0500 cm For constant-rate drying in a batch dryer, all particles have the same
Heat-transfer rate: residence time. From pilot-plant data, u ¼ 8 min for drying.
For the large, continuous operation, (18-91) applies, with X ds ¼
ðCP Þs ¼ 0:20 Btu=lb- F; T y ¼ 145 F ¼ T go ¼ T ds ; 0:05 and Xo ¼ 0.20. Thus,
ms ¼ 10;000=ð1 þ 0:2Þ ¼ 8;330 lb=h dry sand; and 2 3
8
DH vap 1 exp
y ¼ 1;011 Btu=lb; T ws ¼ 70 F: 6 t 7
0:05 ¼ 0:20641
7
5
From (18-69), ð8=tÞ

Q ¼ 8;330f0:20ð145 70Þ þ 0:20ð1Þð145 70Þ

þð0:20 0:05Þð1;011Þg ¼ 1;510;000 Btu=h Solving this nonlinear equation, t ¼ 13:2 minutes average residence
time for particles. Only
Air rate: ð0:20 0:05Þ
ð8Þ ¼ 6 min
ð0:20 0:0Þ
1;510;000
mg ¼
½ð0:242Þ þ ð0:01Þð0:447Þ ð1;000 145Þ residence time would be required in a batch dryer to dry to 5% mois-
¼ 7;170 lb=h dry air ture. Therefore, more than double the residence time is needed in
the continuous unit.
770 Chapter 18 Drying of Solids

Bed height: [32], drying may be needed to preserve required properties

and maintain activity of bioproducts. If a proper drying
To achieve the average residence time of 13.2 minutes ¼ 0.22 h, the
expanded-bed volume, and corresponding bed height, must be method is not selected or adequately designed, the bioprod-
uct may degrade during dewatering or exposure to elevated
ms t 8;330ð0:22Þ
Vb ¼ ¼ ¼ 27:8 ft3 temperatures. For example, the bioproduct may be subject to
rb 66
oxidation and thus require drying in a vacuum or in the pres-
Vb 27:8ð4Þ ence of an inert gas. It may degrade or be contaminated in the
Hb ¼ ¼ ¼ 1:6 ft
pD2 =4 3:14ð4:7Þ2 presence of metallic particles, requiring a dryer constructed
of polished stainless steel. Enzymes may require pH control
during drying to prevent destabilization. Some bioproducts
may require gentle handling during the drying process.
Of major concern is the fact that many bioproducts are
§18.6 DRYING OF BIOPRODUCTS
thermolabile, in that they are subject to destruction, decom-
The selection of a dryer is often a critical step in the design of position, or great change by moderate heating. Table 18.7
a process for the manufacture of a bioproduct. As discussed lists several examples of bioproduct degradation that can
in several chapters of the Handbook of Industrial Drying occur during drying at elevated temperatures. As shown, the

Table 18.7 Examples of Degradation of Bioproducts at Elevated Temperatures

Product Type of Reaction Degradation Processes Result

Live microorganisms Microbiological changes Destruction of cell membranes Denaturation of protein

Death of cells
Lipids Enzymatic reactions Peroxidation of lipids Reaction with other components
(discoloration of the product) (including proteins and
vitamins)
Proteins Enzymatic and chemical Total destruction of amino acids Denaturation of proteins and
reactions enzymes
Derivation of some individual amino acids Partial denaturation, loss of
nutritive value
Cross-linking reaction between amino acids Change of protein functionality
Enzyme reaction
Polymer carbohydrates Chemical reactions Gelatination of starch Improved digestibility and
energy utilization
Hydrolysis Fragmentation of molecule
Vitamins Chemical reactions Derivation of some amino acids Partial inactivation
Simple sugars Physical changes Caramelization Loss of color and flavor
(Maillard-Browning reaction)
Melting

Source: Handbook of Industrial Drying [32]

Table 18.8 Selection of Dryer for Representative Bioprocesses

Bioproduct Dryer Type Comments

Citric acid Fluidized-bed dryer Feed is wet cake from a rotary vacuum filter
Pyruvic acid Fluidized-bed dryer Feed is wet cake from a rotary vacuum filter
L-Lysine (amino acid) Spray dryer Feed is solution from an evaporator
Riboflavin (Vitamin B2) Spray dryer Feed is solution from a decanter
a-Cyclodextrin (polysaccharide) Fluidized-bed dryer Feed is wet cake from a rotary vacuum filter
Penicillin V (acid) Fluidized-bed dryer Feed is a wet cake from a basket centrifuge
Recombinant human serum albumin (protein) Freeze-dryer Feed is from sterile filtration
Recombinant human insulin (protein) Freeze-dryer Feed is wet cake from a basket centrifuge
Monoclonal antibody (cell) No dryer Product is a phosphate-buffered saline (PBS) solution
a-1-Antitrypsin (protein) No dryer Product is a PBS solution
Plasmid DNA (parasitic DNA) No dryer Product is a PBS solution

Figure 18.39 Price and production volume of representative
bioproducts [35].
result of such exposure is serious and unacceptable. To
avoid such degradation, many bioproducts are dried at
near-ambient or cryogenic temperatures. The most widely
used dryers for sensitive bioproducts, particularly solutions
of enzymes and other proteins, are spray dryers and freeze-
dryers (i.e., lyophilizers) [33, 34].
Heinzle et al. [35] consider dryer selection for 11 different
bioprocesses, as listed in Table 18.8. The bioproducts cover
more than a seven-fold range of product value and more than
a six-fold range of annual production rate, as shown in Figure
18.39. It is interesting to note that the three most expensive
SUMMARY
1. Drying is the removal of moisture (water or another
volatile liquid) from wet solids, solutions, slurries, and
pastes.
2. The two most common modes of drying are direct, by
heat transfer from a hot gas, and indirect, by heat trans-
fer from a hot wall. The hot gas is frequently air, but can
be combustion gas, steam, nitrogen, or any other non-
reactive gas.
3. Industrial drying equipment can be classified by opera-
tion (batch or continuous), mode (direct or indirect), or
the degree to which the material being dried is agitated.
Batch dryers include tray dryers and agitated dryers.
Continuous dryers include: tunnel, belt or band, turbo-
tray tower, rotary, screw-conveyor, fluidized-bed,
spouted-bed, pneumatic-conveyor, spray, and drum. Dry-
ing can also be accomplished with electric heaters, infra-
red radiation, radio frequency and microwave radiation,
and also from the frozen state by freeze-drying.
4. Psychrometry, which deals with the properties of air–
water mixtures and other gas–moisture systems, is use-
ful for making drying calculations. Psychrometric
Summary 771
bioproducts are not dried, but produced as phosphate-
buffered saline solutions. The least-expensive and highest-
volume bioproducts use either fluidized-bed or spray dryers.
The fluidized-bed dryers are used with relatively stable bio-
molecules, and operate at near-ambient temperatures. The
two bioproducts at intermediate levels of price and volume
use freeze-dryers.
Intermittent Drying of Bioproducts
As discussed in §13.8, batch-distillation operations can be
improved by controlling the reflux ratio. Similarly, batch-
drying operations can be improved, particularly for heat-sen-
sitive bioproducts, by varying conditions during the drying
operation. This technique is referred to as intermittent dry-
ing. Although the concept has been known for decades, it is
only in recent years that it has received wide attention, as dis-
cussed by Chua et al. [36]. The intermittent supply of heat is
beneficial for materials that begin drying in a constant-rate
period, but dry primarily in the falling-rate period, where the
rate of drying is controlled by internal heat and mass transfer.
In traditional drying, the external conditions are constant and
the surface temperature of the material being dried can rise to
unacceptable levels. In intermittent drying, the external con-
ditions are altered so that the surface temperature does not
exceed a limiting value. In the simplest case, the heat input
to the material is reduced to zero during a so-called temper-
ing phase, while interior moisture moves to the surface so
that a constant-rate period can be resumed. The benefits of
intermittent drying have been demonstrated for a number of
products, including grains, potatoes, guavas, bananas, car-
rots, rice, corn, clay, cranberries, apples, peanuts, pineapples,
sugar, beans, ascorbic acid, and b-carotene.
(humidity) charts are used for obtaining the temperature
at which surface moisture evaporates.
5. For the air–water system, the adiabatic-saturation tem-
perature and the wet-bulb temperature are, by coinci-
dence, almost identical. Thus, surface moisture is
evaporated at the wet-bulb temperature. This greatly
simplifies drying calculations.
6. Most wet solids can be grouped into one of two categories.
Granular or crystalline solids that hold moisture in open
pores between particles can be dried to very low moisture
contents. Fibrous, amorphous, and gel-like materials that
dissolve moisture or trap it in fibers or very fine pores can
be dried to low moisture contents only with a gas of low
humidity. The second category of materials can exhibit a
significant equilibrium-moisture content that depends on
temperature, pressure, and humidity of the gas.
7. For drying calculations, moisture content of a solid and a
gas is usually based on the bone-dry solid and bone-dry
gas. The bound-moisture content of a material in contact
with a gas is the equilibrium-moisture content when the
gas is saturated with the moisture. The excess-moisture
772 Chapter 18 Drying of Solids
content is the unbound-moisture content. When a gas is
not saturated, excess moisture above the equilibrium-
moisture content is the free-moisture content. Solid
materials that can contain bound moisture are hygro-
scopic. Bound moisture can be held chemically as water
of hydration.
8. Drying by direct heat often takes place in four periods.
The first is a preheat period accompanied by a rise in
temperature but with little moisture removal. This is fol-
lowed by a constant-rate period, during which surface
moisture is evaporated at the wet-bulb temperature. This
moisture may be originally on the surface or moisture
brought rapidly to the surface by diffusion or capillary
action. The third period is a falling-rate period, during
which the rate of drying decreases linearly with time
with little change in temperature. A fourth period may
occur when the rate of drying falls off exponentially
with time and the temperature rises.
9. Drying rate in the constant-rate period is governed by the
rate of heat transfer from the gas to the surface of the
solid. Empirical expressions for the heat-transfer co-
efficient are available for different types of direct-heat
dryers.
10. The drying rate in the falling-rate period can be deter-
mined by using empirical expressions with experimental
REFERENCES
1. Handbook of Industrial Drying, 2nd ed., A.S. Mujumdar, Ed., Marcel
Dekker, New York (1995).
2. Perry’s Chemical Engineers’ Handbook, 8th ed., D.W. Green and R.H.
Perry, Eds., McGraw-Hill, New York (2008).
3. Walas, S.M., Chemical Process Equipment, Butterworths, Boston
(1988).
4. van’t Land, C.M., Industrial Drying Equipment, Marcel Dekker, New
York (1991).
5. Uhl, V.W., and W.L. Root, Chem. Eng. Progress, 58, 37–44 (1962).
6. McCormick, P.Y., in Encyclopedia of Chemical Technology, 4th ed.,
John Wiley & Sons, New York, Vol. 8, pp. 475–519 (1993).
7. Keey, R.B., Introduction to Industrial Drying Operations, Pergamon
Press, Oxford (1978).
8. Lewis, W.K., Mech. Eng., 44, 445–446 (1922).
9. Faust, A.S., L.A. Wenzel, C.W. Clump, L. Maus, and L.B. Anderson,
Principles of Unit Operations, John Wiley & Sons, New York (1960).
10. Luikov, A.V., Heat and Mass Transfer in Capillary-Porous Bodies,
Pergamon Press, London (1966).
11. Sherwood, T.K., Ind. Eng. Chem., 21, 12–16 (1929).
12. Sherwood, T.K., Ind. Eng. Chem., 21, 976–980 (1929).
13. Marshall, W.R., Jr., and O.A. Hougen, Trans. AIChE, 38, 91–121
(1942).
14. Gamson, B.W., G. Thodos, and O.A. Hougen, Trans. AIChE, 39, 1–35
(1943).
15. Wilke, C.R., and O.A. Hougen, Trans. AIChE, 41, 445–451 (1945).
16. Hougen, O.A., H.J. McCauley, and W.R. Marshall, Jr., Trans.
AIChE, 36, 183–209 (1940).
data. Diffusion theory can be applied in some cases
when moisture diffusivity is available or can be
measured.
11. For direct-heat dryer models, material and energy bal-
ances are used to determine rates of heat transfer from
the gas to the wet solid, and the gas flow rate.
12. A useful model for a two-zone belt dryer with through-
circulation describes the changes in solids-moisture con-
tent both vertically through the bed and in the direction
of belt travel.
13. A model for preliminary sizing of a direct-heat rotary
dryer is based on the use of a volumetric heat-transfer
coefficient, assuming that the gas flows through curtains
of cascading solids.
14. A model for sizing a large fluidized-bed dryer is based
on the assumption of perfect solids mixing in the dryer
when operating in the bubbling-fluidization regime. The
procedure involves taking drying-time data from batch
operation of a laboratory fluidized-bed dryer and correct-
ing it for the expected solid-particle-residence-time dis-
tribution in the large dryer.
15. Many bioproducts are thermolabile and thus require
careful selection of a suitable dryer. Most popular are
fluidized-bed dryers, spray dryers, and freeze-dryers.
17. Carslaw, H.S., and J.C. Jaeger, Heat Conduction in Solids, 2nd ed.,
Oxford University Press, London (1959).
18. Newman, A.B., Trans. AIChE, 27, 310–333 (1931).
19. Sherwood, T.K., Ind. Eng. Chem., 24, 307–310 (1932).
20. Gilliland, E.R., and T.K. Sherwood, Ind. Eng. Chem., 25, 1134–1136
(1933).
21. Walker, W.H., W.K. Lewis, W.H. McAdams, and E.R. Gilliland, Princi-
ples of Chemical Engineering, 3rd ed., McGraw-Hill, New York (1937).
22. Ceaglske, N.H., and F.C. Kiesling, Trans. AIChE, 36, 211–225 (1940).
23. Keey, R.B., Drying Principles and Practice, Pergamon Press, Oxford
(1972).
24. Genskow, L.R., Ed., Scale-Up of Dryers, in Drying Technology,
12(1, 2), 1–416 (1994).
25. Thygeson, J.R., Jr., and E.D. Grossmann, AIChE Journal, 16, 749–754
(1970).
26. Matchett, A.J., and M.S. Sheikh, Trans. Inst. Chem. Engrs., 68, Part A,
139–148 (1990).
27. McCormick, P.Y., Chem. Eng. Progress, 58(6), 57–61 (1962).
28. Schofield, F.R., and P.G. Glikin, Trans. Inst. Chem. Engrs., 40, 183–
190 (1962).
29. Langrish, T.A.G., R.E. Bahu, and D. Reay, Trans. Inst. Chem. Engrs.,
69, Part A, 417–424 (1991).
30. Ergun, S., Chem. Eng. Progr., 48, (2), 89–94 (1952).
31. Fogler, H.S., Elements of Chemical Reaction Engineering, 3rd ed.
Prentice-Hall, Upper Saddle River, NJ (1999).
32. Handbook of Industrial Drying, 3rd ed., A.S. Mujumdar, Ed., Taylor
and Francis, Boca Raton, FL (2007).

33. Afdull-Fattah, A.M., D.S. Kalonia, and M.J. Pikal, J. of Pharmaceuti-
cal Sciences, 96(8) 1886–1916 (2007).
34. Tang, X., and M.J. Pikal, Pharmaceutical Research, 21(2) 191–200
(2004).
STUDY QUESTIONS
18.1. What are the most commonly employed modes of heat
transfer for drying? Does the temperature of the solid during drying
depend on the mode?
18.2. Why is there such a large variety of drying equipment?
18.3. What is the difference between a direct-heat dryer and an
indirect-heat dryer?
18.4. For what types of wet solids can fluidized-bed, spouted-
bed, and pneumatic-conveyor dryers be used?
18.5. What is freeze-drying and when is it a good choice?
18.6. What is psychrometry?
18.7. What are the differences among absolute humidity, relative
humidity, and percentage humidity?
18.8. What is the wet-bulb temperature? How is it measured?
How does it differ from the dry-bulb temperature?
18.9. What is the adiabatic-saturation temperature? Why is it
almost identical to the wet-bulb temperature for the air–water
system, but not for other systems?
EXERCISES
Section 18.1 (Use of the Internet is encouraged for the exercises of
this section.)
18.1. Continuous dryer selection.
The surface moisture of 0.5-mm average particle size NaCl
crystals is to be removed in a continuous, direct-heat dryer with-
out a significant change in particle size. What types of dryers
would be suitable? How high could the gas feed temperature
be?
18.2. Batch-dryer selection.
A batch dryer is to be selected to dry 100 kg/h of a toxic, tempera-
ture-sensitive material (maximum of 50 C) of an average particle size
of 350 mm. What dryers are suitable?
18.3. Dryer selection for a milky liquid.
A thin, milk-like liquid is to be dried to produce a fine powder.
What types of continuous, direct-heat dryers would be suitable? The
material should not be heated above 200 C.
18.4. Dryer selection for different feeds.
The selection of a batch or continuous dryer is determined
largely by feed condition, temperature-sensitivity of the mate-
rial, and the form of the dried product. Select types of batch
and continuous dryers that would be suitable for the following
cases: (a) A temperature-insensitive paste that must be main-
tained in slab form. (b) A temperature-insensitive paste that can
be extruded. (c) A temperature-insensitive slurry. (d) A thin
liquid from which flakes are to be produced. (e) Pieces of lum-
ber. (f) Pieces of pottery. (g) Temperature-insensitive inorganic
crystals for which particle size is to be maintained and only sur-
face moisture is to be removed. (h) Orange juice to produce a
powder.
18.5. Solar drying for organic materials.
Solar drying has been used for centuries to dry, and thus pre-
serve, fish, fruit, meat, plants, spices, seeds, and wood. What are the
Exercises 773
35. Heinzle, E., A.P. Biwer, and C.L. Cooney, Development of Sustainable
Bioprocesses, John Wiley & Sons, England (2006).
36. Chua, K.J., A.S. Mujumdar, and S.K. Chou, Bioresource Technology,
90, 285–295 (2003).
18.10. Under what drying conditions is the moisture-evaporation
temperature equal to the wet-bulb temperature?
18.11. Distinguish among total-moisture content, free-moisture
content, equilibrium-moisture content, unbound moisture, and
bound moisture.
18.12. What are the different periods that may occur during a dry-
ing operation and under what conditions do they occur?
18.13. What is the critical moisture content?
18.14. What are the two most applied theories to the falling-rate
drying period?
18.15. In the dryer models for a belt dryer with through-
circulation and a direct-heat rotary dryer, is the rate of drying based
on heat transfer or mass transfer? Why?
18.16. What are the regimes of fluidization of a bed of particles
by a gas? What regime of operation is preferred for drying?
18.17. When selecting a dryer type, why do bioproducts require
special considerations?
advantages and disadvantages of this type of drying? What other
types of dryers can be used to dry such materials? What type of
dryer would you select to continuously dry beans?
18.6. Advantages of fluidized-bed dryers.
Fluidized-bed dryers are used to dry a variety of vegetables,
including potato granules, peas, diced carrots, and onion flakes.
What are the advantages of this type of dryer for these types of
materials?
18.7. Production of powdered milk.
Powdered milk can be produced from liquid milk in a three-stage
process: (1) vacuum evaporation in a falling-film evaporator to a
high-viscosity liquid of less than 50 wt% water; (2) spray drying to
7 wt% moisture; and (3) fluidized-bed drying to 3.6 wt% moisture.
Give reasons why this three-stage process is preferable to a single-
stage process involving just spray drying.
18.8. Drying pharmaceutical products.
Deterioration must be strictly avoided when drying pharmaceuti-
cal products. Furthermore, such products are often produced from a
nonaqueous solvent such as ethanol, methanol, acetone, etc. Explain
why a closed-cycle spray dryer using nitrogen is frequently a good
choice of dryer.
18.9. Drying of paper.
Paper is made from a suspension of fibers in water. The process
begins by draining the fibers to a water-to-fiber ratio of 6:1, fol-
lowed by pressing to a 2:1 ratio. What type of dryer could then be
used to dry a continuous sheet to an equilibrium-moisture content of
8 wt% (dry basis)?
18.10. Importance of drying green wood.
Green wood contains from 40 to 110 wt% moisture (dry basis)
and must be dried before use to just under its equilibrium-moisture
content when in the final environment. This moisture content is usu-
ally in the range from 6 to 15 wt% (dry basis). Why is it important to