ResearchGate

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/293304987

SIMPLE PROCEDURE FOR DESIGN OF A SPRAY DRYER.

Artícle • June 1977

CITATIONS READS
0 3,511

A. S. Mujumdar Vijay M. Jog

McGill University Carleton University
833 PUBLICATIONS 15,866 CITATIONS 68 PUBLICATIONS 885 CITATIONS

SEE PROFILE SEE PROFILE

2 authors:
Some of the authors of this publícatíon are also workíng on these related projects:

project dehydration of veggie ingredients View project

All content following this page was uploaded by Vijay M. Jog on 09 July 2016.

Food Resources and Utilization View project

The user has requested enhancement of the downloaded file.

 »ROBLEM S
Reprinted from the Journal of the Institution of Engineers (India), vol 57 pt CH 3 June, 1977
Design a sp
[ollowing infc

Moisture r
UDC 664.8.037 Solids in f
Ambient i
A Simple Procedure for Design of a Spray Dryer Ambient
A S Mujumdar, Non-member V Jog, Non-member Resideno
Powder i
Specific
Volume
A simple procedure is proposed for design of a spray dryer. The procedure is tested by comparing the Inlet tei
design calculations with data supplied by a large commercial installation for spray drying of skim milk. Outlet 1
Despite numerous simplifying assumptions, the agreement between the proposed design and the opera- peed tt
ting dryer was found to be very good.
\ ■Water
peed i
NOTATIONS INTRODUCUON Solids Oil-
fo ¡ Heat
A = latent heat Spray drying converts a liquid feed into dry powder by
= heat capacity of dry solid Cft = atomizing the feed into a stream of hot air contained in a j The
1
chamber of suitable geometry. Uniform drying, consistency tion oi
heat capacity of moisture Cs = heat
of physical properties, uniformity of size and shape are some
capacity of humid air Dvm = volume of the notable advantages of this suspended-air drying
mean diameter d — technique. Evaporative cooling coupled with short residence
times in the chamber avoid scorching of heat-sensitive Th
diameter of rotary wheel e
products. Masters1 has sum- marized the vast body of
G = mass flow rate literature on spray drying. vessel
H = enthalpy dryin
h = height of the vane A very wide variety of possible combinations of chamber
K = constant M — mass flow rate N geometries and atomization techniques makes it impossible
to develop a generally valid design procedure. This, coupled (í
— rpm n = number of vanes Q = with the various types of heat sources available, relative
quantity of heat QL = heat loss from directions of spray and drying air flows, physical and
the dryer T = temperature UH = Chemical characteristics of the product being dried,
characteristics of the final product, etc., makes it mandatory
humid volume of air W = moisture
to conduct pilot plant studies before designing a full-scale
content unit.

SUBSCRIPTS
The objective of this paper is to ¡Ilústrate that, despite the
numerous simplifying assumptions explicit or implicit in the
proposed procedure, it is possible to make reason- able
design calculations1.

a = air
s = solid In this paper the design procedure is developed and
W = water illustrated for the case of spray drying skim milk-a process
1 = inlet used almost universally for this purpose. The input data were
2 = outlet supplied by one of the world’s leading spray dryer t ti
wb = wet bulb |
manufacturers. The design valúes are then compared with
0 = ambient physical data for an operational unit.
4

A S Mujumdar and V Jog are with McGill University, Montreal, Cañada

This paper was received on June 17,1976. Written discussion on this paper will be received until August 31,1977.

134 IE (I) Journal-CH

PROBLEM STATEMENT where
Design a spray drying system for skim milk, using the
following information:

Moisture in powder 3.5%

Solids in feed 48%
Ambient relative humidity 10%
Ambient temperature 10°C
Residence time in dryer 30 sec
Powder rate 1 200 kg/hr
Specific heat of solid 0.4 kcal/kg
Volume mean particle diameter 100 m
Inlet temperature ( air ) 215°C
Outlet temperature ( air ) 95°C
Feed temperature 60°C
Water evaporation rate 1 212 kg/hr
Feed rate 2 413 kg/hr
Solids rate 1 158 kg/hr
Oil-fired heater for air heating
10%
Heat losses V max

The feed outlet temperature is 80°C and 100% collec-

tion of the product takes place in the cyclones. = Ü_____?“Üx 100 x 100
% Ti Tamfj
The drying chamber is a flat-bottomed cylindrical vessel.
Conical bottom chambers are also used in spray drying skim
milk. '
PROBLEM STATEMENT where
QSl = Cds AT + fVSl cw AT (3) hard, overdry particles). For ware articles and for pressing
use in treatment of earthen- in the tile and crockery
Qa = C,( AT) + HX (4)
Here Qs¡ and Qa are enthalpies of feed and air res- Fo/57, June 1977 135
pectively.
Qa, = 52.0 kcal/kg, Qa% = 23 + 596 H2 Qs, — 89.0 kcal/kg,
QSi — 35 kcal/kg
Substituting these valúes in equations (1) and (2),
two simultaneous equations for Ga and H.¿ are obtained.
Ga = 28 282 kg/hr and
H2 = 0.044 kg/kg of dry air at 95°C.
Heat load = 1.58 x 10® kcal/hr, from equation (2)
If the operation is adiabatic (ie there is no heat loss ), then Ga
= 22 843 kg/hr and H2 =0.053 kg/kg of dry air.

DRYER EFFICIENCY
MÁXIMUM THERMAL EFFICIENCY ( IDEAL CASE )
Heat used in evaporation (exhaust air is saturated)
Heat input
/ 215 - 69.4 \
\ 215 — 10 /
72%
OVERALL THERMAL EFFICIENCY
(M8)(Ts)i(aji(Ws)i Heat used in evaporation (for adiabatic
Voverall
Heat input operation)

/ 215 - 98.3 \ VX 100

215 - 10 /
56.9%

EVAPORATIVE EFFICIENCY
Actual evaporative capacity
Veva Capacity obtainable with saturated exhaust air
p
/ 215 - 98.3 \ \
100 X 100
215 - 69.4 / =
80.2%

COMPARATIVE EVALUATION OF ATOMIZERS

Fig 1 Dayer diagram for calculation of heat and mass balance

MOISTURE AND HEAT Different types of

BALANCE atomizers are used
commercially to dry the same
Moisture balance product, depending upon the
MsiWs, - WsQ end use of the product. For
example, both pressure and
= G a (H t - H x ) centrifugal (spinning) disc
atomizers are used in the
ceramic industry to dry
Heat balance kaolín. In the production of
G 0 Qo, + M s Q s , = GaQai paper coating powders a
centrifugal atomizer produces
+ M S Q S + Q 2 (2) the required, fine, uniform
powder of kaolin (free from
industry, pressure nozzles yield a granular, dust-free The design procedure consists simply of choosing
product of uniform and constant moisture content and high reasonable valúes for d, n and h which will result in
bulk density. Thus there is no blanket method of selection of 1. 9 < Mp < 5.4.
atomizer, depending only on the feed material although the
physical ( and sometimes Chemical) charac- teristics of the The following set of design parameters is one of the
feed also influence the choice of the atomizer type. numerous possible combinations suitable for the specified
duty.

Extensive empirica) as well as theoretical work has been Wheel diameter = 22.9 cm; Wheel speed = 15 000 rpm
done on design of atomizers. Friedman, Gluckest, Marshall 2, Number of vanes = 20 ; Height of vanes = 1.9 cm
Herring and Marshall3, Frazer,4 Masters and Mohtadi,5 and
Scout, Robinson, et al6 have deve- loped correlations for the DESIGN OF SPRAY DRYING CHAMBER
design of atomizer wheels or dises. Nozzle atomizers have
been studied by Marshall,7 Green,8 Doumas and Laster,9 A concurrent air flow pattern is recommended because of
Dombrowski and Hassen20 and Moulton and Turner.11 the heat sensitivily of milk. This allows higher inlet
Droplet size predictions can be made with the help of the temperature to be used at inlet to speed up the drying without
classical Lewis- Nukiyama-Tanasawa equation, or the Tate- damaging the product.
Marshall equation12.
Two relations have been proposed in the literature to
For spray drying milk, rotary atomizers are used most calcúlate the spray trajectofy from a rotary atomizer, which
frequently. Some of the advantages of rotary atomizers are: form the basis for selecting a suitable chamber diameter:
(1) Greater realiability and flexibility;
(2) Little wear of atomizing disc; atomizing charac-
teristics not affected by possible wear of the disc;
(3) Wide capacity range without change in atomizing (1) Frazer, Eisenklam, Dombrowski4
characteristics; „ „„ ¿°'21 M°-2
R max 7.48 —---------
(4) Little risk of clogging; jyo.i6 (7)
(5) Ease of automation and control.
where Rmax is the radial distance in cm at which 99 % of the
spray falls 0.91 meters below atomizer

Taking these into account, a rotary atomizer with radial (2) Herring and Marshall3
vanes is recommended for this design. ¿0.2
R 11.87 M0.25
DESIGN OF ROTARY ATOMIZER , jyo.ia
where R99 is radial distance in cm which ineludes 99 % of
For industrial rotary atomizer Herring-Marshall equation3 the mass of spray. For design under consideration, Rmaz =
can be used to predict parameter valúes. The basic 1.83 m and R = 3.74 m.
99
parameters aie the droplet size Dvm microns, feed rate ML
kg/min, height of vanes h cm, rotation speed N rpm, diameter The substantial difference between the two correlations
of wheel d cm, number of vanes n. The equation is given as: may be attributed to the following factors:

z>«. = (5)
(1) Both correlations ignore the effeets of wall and
(Nd)0-**(nh) °-12 W
therefore give only a conservative estímate of
The constant K for industrial dryers of large capacity is about droplet penetration;
29.4 x 104. This correlation is valid over the following ranges
of parameters: (2) The radial travel of droplets also depends upon
factors like wheel design and speed, feed rate, air
Variable disperser design and its location. These are not
Wheel diameter (cm) considered in these correlations.
Wheel speed (rpm) Range of Operations
Mp = liquid loading on vane 19.0 - 23.0 10 000 - 18
With the limited information available, only a conser-
000
vative judgement can be made of the radial distance, by
where 0.9 - 5.4 taking the arithmetic average of the two valúes to size the
Mp = kg/min om (6) drying chamber which yields R =2.75 m or chamber
diameter of 5.5 m. In practice, pilot scale data are needed for
a confident specification of the chamber size.

136 IE (1) Journal-CH

HEIGHT OF DRYING that is, particles of 200-600 conditions. The chamber diameter and
CHAMBERS m. For high degree of height as computed by the
atomization ( 5-20 m The demand hp then can highly simplified procedure
Drying chambers are be estimated empirically,
designed to handle an air particles), the power required outlined here are in
volume containing suíficient goes up to 1.1 hp/kg of feed hp = vol flow rate at outlet remarkably cióse agreement
heat for drying the spray per min. For the present case, (m3/min ) x ~ x system with those of the actual
droplets and to provide an air power consumption for K.commercial unit. The latter
residence time sufficient for atomization is estimated to r information is, however,
droplets to be dried to be 8.2 hp. esistance (in cm of water) proprietory.
particles of desired moisture where K = 27.5 for most
SUMMARY OF DESIGN commercial dryers14. This Despite this agreement,
contení. Residence times pilot scale tests are
ranging from 3 to 50 sec are yields a fan power
(i) Inlet Air Flow Rate = 28 requirement of about 475 hp recommen- ded before
found in commercial dryers. 282.0 kg/hr designing a full-scale unit.
For the present case the for air volume flow of 524
3 Location and design of inlet
residence time is specified to m /min at 25 cm of water
and outlet ducts can modify
be 30 sec. pressure drop.
the flow field and henee the
Drying rate curves for a (ii) Total heat input = 1.58 ATOMIZER POWER droplet trajectories and
droplet can give information x 106 kcal/hr Heat CONSUMPTION drying kinetics very
on mínimum residence time loss significantly. A more
needed in drying, but due to kcal/hr According to a leading fundamental approach is
the difference in evaporation Heat required/kg of spray dryer manufactuer, a therefore necessary before
characteristics of single droplet product valué of 0.22 hp/kg of feed reliable generalized design
and sprays, residence time kcal per min gives a fairly procedures and criteria can
requirements are based on accurate estímate of the be^ established. Some
pilot plant tests. For many Estimated Efficiencies atomizer power consumption attempts in this direction ha
producís dried in concurrent Maximim thermal foi applications that require a ve begun to appear. 15,16

flow dryers with rotary air efñciency = low degree of atomization, However, empiricism will
flow, residence times range 71.00% Overall continué to be the basis of
from 20 to 40 sec. thermal efñciency spray dryer designs for many
= 57 % years to come.
HEIGHT OF DRYER Evaporative
efñciency CONCLUDING REMARES
%
Various theoretical
Volumetric flow rate Atomizer Design approaches for designing a
Wheel dia meter spray dryer
= 22.9 cm = 20 system have
V= = 8-73 m3/sec Number of vanes = appeared
1.90 cm =in15the literature, but
3 600 ' the design
Height of vanes 000 rpm = 8.2is still based on
Rotational speed hpempirical methods. There is
Average velocity of dry air no technique available for
Power requirement
measuring drying kinetics
v = — V --- 0.369 m/sec ©ver time scales of a few
irí/2 Residence time = 30 sec
seconds; thus all theoretieal
Height required Diameter of chamber = 5.5 m Height
models
v x residence time = 11 m (for a 30 sec residence time) of the chamber = 11.0 m
assume a constant drying rate similar materials. Such
which ultimately leads to a information is almost
POWER REQUIRED FOR AIR MOVING dryer which is too small. invariably pro- prietory and
Parameters
EQUIPMENT AND ATOMIZERS henee not easily available.
requirement = 475 hp Until techniques are
developed for independent REFERENCES
To calcúlate the required water due to use of wet measurement of spray drying
horsepower, the system resis- scrubber are reasonable kinetics, the design of these 2. K Masters. Spray Drying.
tance to gas flow must be estimates for industrial units. dryers will continué to be Leonard-Hill Publishing Co,
known. A total resistance of If total product recovery is by based on scale-up from the 1972.
25 cm of water at operating the use of a filter type biggest pilot plant one can 3. S J Friedman, F A
conditions for the drying collector, the total system afford, using whatever Gluckert and W R Marshall.
chamber with cyclone resistance will be 20-25 cm experience one has Chemical Engineering
collectors and 20-25 cm of of water at operating accumulated in the past on Progress, vol 48, no 4, 1952,

Vol 57, June 1977 137

 p 181. 1975, p 793.
4. W M Herring and W R
Marshall. American Instituíe
of Chemical Engineering
Journal, vol 1, no 2, 1955,
p 200.
5. R P Frazer, et al. Briíish
Chemical Engineering, vol 2,
no 9, 1957, p 196.
6. K Masters and M F
Mohtadi. Briíish Chemical
Engineering, vol 12, no 12,
1967, p 1890.
7. M N Scott, et al. Journal
of Pharmaceutical Science,
vol 53, no 6, 1964, p 670.
8. W R Marshall. Chemical
Engineering Progress Mono-
graph Series, no 2, 1954, p
50.
9. H L Green. Flow
Properties of Disperse
Systems, North Holland
Publishing Co, Amsterdam,
1957.

10. M Do urnas and R

Laster. Chemical
Engineering Progress, vol
49, no 10, 1953, p 518.
11. N Dombrowski and D
Hasson. American Institute
of Chemical Engineering
Journal, vol 15, no 4, 1969, p
604.
12. G M Tumer and R W
Moulton. Chemical
Engineering Progress, vol
49, no 4, 1953, p 169.
13. R W Tate and W R
Marshall. Chemical
Engineering Progress, vol
49, no 4, 1953, p 169.
14. Koch Technical
Bulletin, BSD, vol 5, 1974.
15. S Katta and W H
Gauvin. American Institute
of Chemical Engineering
Journal, vol 21, 1975, p 143.
16. W H Gauvin, F H
Knelman and S Katta.
International Journal
138
Multiphase Flow, vol 1, IE (/) Journal-CH