@ Pergamon

Plh S0038-092X (96) 00130-2

SolarEnergyVol.59, Nos. 4-6, pp. 121-126,1997
© 1997ElsevierScienceLtd
All rights reserved.Printed in Great Britain
0038-092X/97$17.00+ 0.00

DEHYDRATION OF FOOD CROPS USING A SOLAR DRYER WITH

CONVECTIVE HEAT FLOW

AKWASI AYENSU
Department of Physics, University of Cape Coast, Cape Coast, Ghana
Abstract--A solar drying system designed on the principles of convective heat flow was constructed from
local materials (wood, metals and glass sheets) and used to dry food crops (cassava, pepper, okro,
groundnuts, etc.). The solar collector could transfer 118 W m -~ thermal power to the drying air. The
thermal exchanges within the dryer were determined from a psychometric chart. Ambient air at 32°C and
80% relative humidity (RH) could be heated to 45°C at 40% RH for drying. The crops were dried to a
final moisture content of < 14% and were preserved for a period of one year without deterioration. The
low-temperature drying system ensured the viability of the seeds for planting. The drying process can be
represented by an empirical equation of the form M(t)=Mo exp (-kt) or dM/dt= -kM, where M0 is
the initial moisture content, M(t) is the moisture content at time t, and k is the drying constant. Under
identical conditions, a high value of k was correlated with a shorter drying period. The drying process
takes place in two phases: constant rate and falling rate periods, and the drying equation was solved to
predict the total drying time. The mechanisms for the dehydration are the removal of unbound "free"
water in the cell cavities and of "bound" water (water films) trapped within cells or chemically bound
with solids as water of crystallization. © 1997 Elsevier Science Ltd.

1. INTRODUCTION 2. PSYCHROMETRY AND DRYING

PROCESS
D e h y d r a t i o n o f vegetables and other food crops
by the traditional m e t h o d o f open air sun- In Fig. 1 ( A S H R A E , 1963), if ambient air at
drying is not satisfactory, as the products temperature T A = 3 2 ° C and relative humidity
become infested with bacteria and insects and ( R H ) ~b=80% is heated to T p = 4 5 ° C , then ~b
deteriorate rapidly in the high ambient temper- would be reduced to 40%. I f this heated air is
atures and relative humidity. Ayensu (1983) has used to remove moisture f r o m cassava o f
observed that the solar intensity ranges from Mo = 70% (on wet basis) until an equilibrium ~b
550 to 1075 W m -2 with a total incident energy o f 90% is reached, the temperature o f the drying
per day o f 15-30 M J m -2 in G h a n a . Therefore, air will be reduced to T c = 3 3 . 3 ° C and the
solar drying systems must be developed to uti- humidity ratio, w, would change f r o m 0.0248
lize this energy resource to improve food to 0.0296, or Aw = 0.0048. This means that the
preservation. capability o f the heated air for dehydration is
Killock (1966) has reported post-harvest nearly eight times greater than ambient or
losses o f between 12 and 37% for various f o o d unheated air. The drying paths are represented
crops in G h a n a , due mostly to storage at high by A B C for heated air and A D for unheated air.
moisture content. D e h y d r a t i o n as a m e t h o d o f The energy balance equation for the drying
preserving farm p r o d u c e involves the removal process is
o f moisture to prevent the development o f a mwL=ma C ( T p - Tc) (1)
favourable e n v i r o n m e n t for the g r o w t h o f
where rnw is the mass o f water evaporated from
moulds, bacteria and insects that n o r m a l l y the sample and absorbed by drying air o f mass
cause spoilage. I m p r o v e d solar drying would ma; Tp and Tc are the initial and final tem-
facilitate early c r o p harvest, long-term storage, peratures o f the drying air, respectively,
viability o f seed grains and sale o f quality L = 2.43 M J k g - 1 is the latent heat o f vaporiza-
produce. tion for free water at 38°C, and C is the specific
In this article, the variables and mechanisms heat capacity o f air. The mass o f water evapo-
involved in the dehydration processes will be rated was calculated f r o m the equation
discussed. The design parameters calculated
were quantity o f air needed for drying, depth mw = mi(M0 - M f ) / ( l O O - M f ) (2)
o f f o o d bed, height o f chimney, and pressure where Mr is the final or equilibrium moisture
difference across the f o o d bed. content and mi is the initial mass o f the
121
122 A. A y e n s u

40

10 15 20 T A TC 25 Tp
Enthalpy (h) BTU per pound of dry air

Fig. 1. A S H R A E p s y c h r o m e t r i c c h a r t No. 1 for n o r m a l t e m p e r a t u r e range.

as-received sample. Therefore, to dehydrate densities of ambient and heated air. The density
50kg of cassava at M o = 7 0 % to Mf=14%, difference equation can be written as
mi = 50 kg, then rnw = 32.5 kg.
Since Aw = mw/ma= 0.0048, m a ---6.88 x 103 kg,
Ap=p-p'=P(T21 - T f l ) / R (3)
and the corresponding volume of air required where Tp=318 K, TA= 305 K, P = 1.01
for dehydration, Va = maRTc/P = 6.03 x 103 m 3 x105Pa, R = 2 . 9 x 1 0 2 j k g 1 K - 1 and, there-
(at Tc-- 306.3 K, pressure P = 1.01 x 105 Pa, and fore, A p = 4 . 6 8 x 1 0 - 2 k g m -3. If h=-hx+h2=
gas constant per unit mass of dry air R = 2.9 x 3m, the corresponding pressure difference
102 J kg -~ K - l ) . Alternatively, using eqn (1), because of convection is AP=h(Ap)g=
m , = 7 . 5 2 x 103 kg and ga = 6.59 X 10 3 m 3, 1 . 4 0 N m -2, where g is the acceleration due
which agree with the previously calculated to gravity.
values. Hence, the mean value Va=6.31 X If the total estimated time for drying is 72 h
103m 3. From the psychrometric chart, the when the rock storage system is used, then the
quantity of thermal energy required for the air flow rate within the collector would be
dehydration process is 13.5kJkg -1 and the 0.024 m 3 s-l. Suppose the density of the food
specific volume of dry air required for item (cassava) is 1.5 x 103 kg m -3 and the mass
the heating process is 0.93 m 3 kg-1. is 50 kg, then the volume is 0.033 m 3 (excluding
void fraction). The required rate of air flow
across a unit volume of cassava chips is
3. D E S I G N P R I N C I P L E S A N D P A R A M E T E R S
0.024/0.033 s-1, or 4.4 complete air changes in
The convection of air through the sample bed the food bed per minute. The corresponding
is caused by a pressure drop across it resulting specific air flow rate is 0.029m 3 min -1 kg-l;
from the difference between the density of rela- that is, the dryer containing 50 kg of cassava
tively cool ambient air and the warm air inside chips would be ventilated with an air flow at a
the dryer. In Fig. 2, P1, P2, P3 and P4 are the rate of 1.46 m 3 rain- 1.
pressure levels, while p and p' are the respective The depth of the food bed was chosen to be
 Dehydration of food crops 123

P4
Moist Air
/

Chimney

T I
Solar Radiation p°

SokJr
~ Rodiatlon~ . / ~. . . . . . . . . . . . _P2._ [

p ~4 / ~ -~***~ Plenum Chamber

p' h1

P~ Rack Star(xJe System /

Fig. 2. Schematic diagram of fixed bed dryer with solar col-

lector, plenum chamber, drying chamber and chimney. I v.om j

Fig. 3. General view of dryer.

x ~ 0 . 1 m, so that the resistance to air flow structed using local materials of wood, scrap
would be minimum. Since the cross-sectional metals and glass sheets.
area of the bed is 0.33 m 2, the air flow velocity The solar air heater was constructed using a
~ 4 . 4 m min -1. As the pressure gradient across single layer of glass (of transmittance > 0.7 for
the food bed is very small, the air flow velocity wavelengths in the range 0.2-2.0~tm and
is proportional to the pressure gradient as opaque to wavelengths greater than 4.5 lam) as
observed in natural convection; i.e. v = glazing material. Granite was used as absorber
7(Ap/Ax), where ? = 0 . 3 2 m 4 Nm -1 is the resis- medium since it has a high thermal diffusivity
tance of the sample to air flow. of 1.27× 1 0 - 6 m 2 s -I. Polystyrene of 3 0 m m
Let the total solar energy received be thickness was used for insulation since it has a
30 MJ m-2. If the optical efficiency and thermal low thermal conductivity of 0.03 W m - a °C- 1.
efficiency of the collector glazing material is 85 The structural frame and panels were built from
and 50%, respectively, then the overall opto- well-seasoned odum boards which could with-
thermal efficiency of the single glazing solar stand termite attack. To minimize atmospheric
collector r/~ 30%. The total thermal energy pro- corrosion, the chimney was constructed from
vided by a unit area of the collector during the galvanized iron sheets. The granite chips were
drying period (3 days) is 3 x 0.3 x 30 MJ m -2. placed on a layer of straw to reduce excessive
The total energy required for dehydration is heat losses to the ground. A galvanized wire
33 x 2.7 M J, which is equivalent to the total mesh was placed at the entrance of the air duct
heat available for drying or 3 x 0.3 x 30 x A, in to keep rodents out.
which case, the calculated collector area The food bed was constructed from a double
A =3.3 m 2. layer of chicken wire mesh with a fairly open
structure to allow drying air to pass through
the food sample, but prevent the pieces of food
items from falling into the plenum chamber.
4. C O N S T R U C T I O N OF DRYER
Access to the drying chamber was via three
Figure 3 shows a general view of the dryer, removable wooden panels made of 1.27 cm ply-
consisting of collector with rock storage system, wood, which overlapped each other to prevent
drying chamber and chimney, which was con- air leakages when closed or inserted. The top
124 A. Ayensu

glazing on the drying chamber provided addi- 100 ' i i , t I ' 1 i ' i '

tional heating and also served as an inspection

• Ambient RH
port. 90
• Renum RH
The chimney was a hollow cylinder of 0.3 m
diameter and 1.90 m height and was constructed 80
from galvanized iron sheets and painted matt
black to obtain elevated temperatures to 7o
improve air flow and increase buoyancy. An
access port was provided for temperature and ~ 60
R H monitoring.
All the wooden components were treated with 50
corroid creosote wood preservative to prevent
termite and fungal attack. The dryer was sited 4o
to face a southerly direction so that an optimum
amount of solar insolation could be collected. 30
The tilt angle of the collector was set at 15 ° at
i I ~ I I i I i I ~ I , I I *
a test site located at 8°N. 2o
8 9 10 11 12 13 14 15 16 17
t (h)
5. C R O P D E H Y D R A T I O N Fig. 5. Diurnal variation of RH for dehydration process.
The diurnal variation of the temperature and
relative humidity of the ambient air and plenum 100 I ' v ' I ' I ' I ' I ' I ' I ' t ' ~ ' ~

chamber are shown, respectively, in Figs. 4 and 90

5. The drying processes were enhanced by the • Solar dehydration curve

heated air at very low humidity. Typical dehy- 80 • Open air sun-drying curve

dration curves for cassava leaves are shown in

7O
Fig. 6. The corresponding dehydration rate is
shown in Fig. 7. The sections of the curves 60
labelled AC represent the constant rate period,
while the section CB is the falling rate period, v
50

as demarcated by the line EF with a correspond-

40
ing critical moisture content Me. The food items
dehydrated in the dryer were superior to those 30

100 I I ' P ' I I I I ' I ' 20

• Absorber temperoture
10
90
• Plenum temperoture
| [ L .~"'-t,,.~A j
• Ambient temperature 0
80 0 20 40 60 80 100 120
t(h)

7O Fig. 6. Dehydration curves for cassava leaves in solar dryer

and open air sun-dryer.
,,%. 6 0
F--
which were sun-dried when evaluated in terms
50 of taste, colour and mould counts. For safe
storage of crops at an R H of 80% 90%, an
40
equilibrium moisture content M e < 1 4 % is
preferred.
30
For thin layer drying, the dehydration curves
shown in Figs. 6 and 7 can be represented by
20 , I , I , I L I i I , I , I , I. i general equations of the form M=Mo exp
8 9 10 11 12 13 14 15 16 17 (-kt) and dM/dt=-kM, respectively. The
t (h) drying constants kc and k e were calculated for
Fig. 4. Diurnal variation of temperature for dehydration the constant rate period and falling rate period.
process. The dehydration parameters for the various
 Dehydration of food crops 125

The collector heat flux can be expressed as

• Solar dehydroUon rate J¢ = C(Te - ra)da/dt (6)

7
• Open air sun-drying rote
where da/dt=A-l(dm/dt) is the ratio of mass
flow rate of the transfer fluid to the collector
area, and C is the specific heat capacity of air.
The parameter da/dt=pA-l(dV/dt), where
d V/dt is the volume flow rate and p is the
density of air. Since dV/dt=O.O24m3s-~,
7 A = 3 . 3 m 2, C = l . 0 2 k J k g - l K -1, p = l . 2 2 k g
m -3 and assuming that F i = 5 5 0 W m -z with
25% atmospheric absorption (Ayensu, 1983),
then J~ = 118 W m - 2 and hence r/= 21%. The
natural convection coefficient for heat trans-
fer for the collector H¢=JJ(Tp--Tc)=9.1
W m 2 K -1.
, t ,~
B i , The amount of sensible heat stored by the
20 40 60 80 100 120 rock pile Qs =msC~(AT)=ps V~C,(AT), where
t(h) p~ = 2.7 x 103 kg m - 3 is density, C~= 794 J
Fig. 7. Dehydration rates for cassava leaves. kg-~ °C-~ is the specific heat capacity,
ms=7.43 x 103 kg m -3 is the mass and V~ is the
volume of rock used as the heat storage
food items are shown in Table 1, where Me is medium, and A T = 7 ° C is the mean daily
the critical moisture content, Mr is the free temperature change associated with the rocks.
moisture content, and Mb is the bound moisture The mean stored energy per unit volume
content. Samples with higher k values will take Q ~ = l . 5 x 107Jm -3 or 4 . 1 7 k W h m -3. Since
a shorter time to be dehydrated. the total storage volume is 2.75m 3, the
Detailed analyses of the dehydration equa- maximum stored energy is l l . 5 k W h or
tions have been given elsewhere (Ayensu and 4.13× 10v J, which would be available to
Asiedu-Bondzie, 1986), where it was shown that enhance drying during the night.
for the constant rate period
[M(t)-M¢]/(Mo-M¢)=exp (-kot¢) (4) 7. CONCLUSIONS

and for the falling rate period ( 1) Based on local weather conditions and ther-
mal processes from psychrometric analysis,
[M(t)-Me]/(M¢-Mc)=exp(-kete). (5) a low-cost, low-temperature and simple to
operate solar dryer was constructed to
6. D I S C U S S I O N dehydrate farm produce.
(2) It took nearly two times longer to dehydrate
The performance of the dryer determines the crops by open air sun-drying compared to
total drying time which depends very much on the solar dryer.
the collector and the thermal exchanges. The (3) An empirical drying equation of the form
overall thermal efficiency of the solar collector M(t)=Moexp(-kt) describes the dehy-
is given as rl=JJFi (Weider, 1982), where Jc is dration process fairly well, and can be used
the heating power collected per unit area and to model and simulate drying of common
F~ is the solar flux intercepted by the collector. tropical farm products.

Table 1. Dehydration parameters of food items

Sample Moisture content (%) Drying period (h) Drying constant ( h - 1)

Mo Mc Me Mr Mb tc t, kc ke

Cassava chips 75 55 14 20 55 12 92 0.031 0.012

Cassava leaves 72 45 14 27 45 10 46 0.047 0.012
Pepper 68 52 14 16 52 10 76 0.025 0.027
Groundnuts 70 47 14 23 47 11 79 0.028 0.016
126 A. Ayensu

8. N O M E N C L A T U R E REFERENCES
Ayensu A., (1983) Analysis and design of a differential
M moisture content sunshine recorder. Solar Energy 31, 5, 517-522.
m mass Killock A. (1966) The Economy o f Ghana. Allen and Union
k drying constant Publications, U.K.
t time ASHRAE (1963) Psychrometric Chart No. 1, Normal Tem-
V volume perature. American Society of Heating and Air-
w humidity ratio conditioning Engineers, U.S.A.
L latent heat of vapourization Ayensu A., Asiedu-Bondzie V., (1986) Solar crop drying
C specific heat capacity with convective self flow and energy storage. Solar Wind
P pressure Technol. 3, 34, 272-278.
Jc heating power collected Weider S. (1982) An Introduction to Solar Energy for Scien-
Fi solar flux tists and Engineers. Wiley, New York.
resistance of food sample to air flow
p density of air
r/ efficiency of collector
relative humidity (RH)