Technical feasibility assessment of a solar chimney for food drying

Text-only Preview

Available online at
Solar Energy 82 (2008) 198–205
Technical feasibility assessment of a solar chimney for food drying
Andre´ G. Ferreira a,*, Cristiana B. Maia b, Ma´rcio F.B. Cortez c, Ramo´n M. Valle c
a Departamento de Cieˆncias Exatas e Tecnologia, Centro Universita´rio de Belo Horizonte, Av. Professor Ma´rio Werneck, 1685. Buritis,
Belo Horizonte, Minas Gerais, CEP 30455-610, Brazil
b Pontif?´cia Universidade Cato´lica de Minas Gerais, Av. Dom Jose´ Gaspar, 500 – Corac¸a˜o Eucar?´stico, Belo Horizonte, Minas Gerais, CEP 30535-901, Brazil
c Departamento de Engenharia Mecaˆnica, Universidade Federal de Minas Gerais, Av. Antoˆnio Carlos, 6627 – Campus Pampulha, Belo Horizonte,
Minas Gerais, CEP 31270-901, Brazil
Received 4 October 2006; received in revised form 6 June 2007; accepted 9 August 2007
Available online 7 September 2007
Communicated by: Associate Editor I Farkas
Solar dryers use free and renewable energy sources, reduce drying losses (as compared to sun drying) and show lower operational
costs than the arti?cial drying, thus presenting an interesting alternative to conventional dryers. This work proposes to study the feasi-
bility of a solar chimney to dry agricultural products. To assess the technical feasibility of this drying device, a prototype solar chimney,
in which the air velocity, temperature and humidity parameters were monitored as a function of the solar incident radiation, was built.
Drying tests of food, based on theoretical and experimental studies, assure the technical feasibility of solar chimneys used as solar dryers
for agricultural products. The constructed chimney generates a hot air?ow with a yearly average rise in temperature (compared to the
ambient air temperature) of 13 ± 1 °C. In the prototype, the yearly average mass ?ow was found to be 1.40 ± 0.08 kg/s, which allowed a
drying capacity of approximately 440 kg.
Ó 2007 Elsevier Ltd. All rights reserved.
Keywords: Solar chimney; Solar drying; Technical feasibility
1. Introduction
and insu?cient drying). Though the arti?cial dryers pro-
vide an improved quality of drying (as they control the
In the near future, the amount of food produced will be
velocity and the temperature of the air?ow), they also con-
insu?cient to feed the world’s population (Mu¨hlbauer
sume a signi?cant amount of energy (fossil or electric) to
et al., 1996). This can be explained by the rapid growth
heat and move the air?ow.
of the world’s population (particularly in developing coun-
To ensure a continuous supply of food for an ever-
tries) as well as by a considerable amount of post-harvest
increasing Brazilian population, and to allow farmers to
losses in foods. To minimize losses, the food materials need
increase their production quality and reduce losses, it is nec-
to be dried to reduce moisture and, in turn, increase their
essary to develop an e?cient drying method with low costs.
shelf life. Natural sun drying requires little investment,
Arti?cial drying is economically feasible, especially when
but has presented signi?cant losses caused by product
used on large farms. Nevertheless, the acquisition and
humidity reabsorption during the rainy period; by contam-
operational costs of these dryers signi?cantly increase the
ination from pathogenic gems, rodents, birds and insects;
costs of the dried product. Therefore, since solar dryers
as well as by enzymatic reactions (caused by heterogeneous
use solar energy (a renewable and low pollutant source of
energy) to dry agricultural products, they in turn present
an interesting and promising alternative. Many solar food
Corresponding author. Tel.: +55 31 3498 0196; fax: +55 31 3378 9294.
E-mail address: [email protected] (A.G. Ferreira).
dryers have been developed over the past few years
0038-092X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
Fig. 1. Schematic representation of the solar chimney prototype as a solar dryer.
(El-Sebaii et al., 2002; Pangavhane et al., 2002; Bena and
with sheets of wood and covered by ?berglass at a diameter
Fuller, 2002; Condor?´ and Saravia, 2003; Ivanova et al.,
of 1.0 m. The cover was made of a plastic thermodi?usor
2003; Simate, 2003; Chen et al., 2005; Forson et al., 2007;
?lm. The cover, with a diameter of 25 m, was set 0.5 m above
Hossain and Bala, 2007; Madhlopa and Ngwalo, 2007).
the ground level, using a metallic structure. The absorber
Fig. 1 shows a sketch of the solar chimney as a dryer device,
ground was built in concrete and painted in black opaque.
which shows some advantages when compared to conven-
The height of the air entrance was reduced to 0.05 m to
tional solar dryers.
minimize the e?ects of the external wind speed under the
The solar chimney is composed of a central tubular
coverage and the consequent cooling of the absorber
tower ?xed to a translucent circular cover, opened at the
ground. For the same purpose, a plastic ?lm was installed
edges (Fig. 1). During the solar radiation incidence period,
around the dryer (at a distance of 2.5 m from the device
a fraction of the incident solar radiation on the cover is
boundaries) to avoid the cooling of the absorber ground.
absorbed by the ground and the drying product, which is
Railcars were used to aid the placement and removal of
then converted into thermal energy. The heat is transferred
the drying products inside the solar chimney, as shown in
by convection through the air and in turn to the product.
Fig. 1.
The hot air?ow enters the tower and creates an updraught
from buoyant forces. The ambient air ?ows from the
periphery to the center of the circular collector. In the pro-
2.2. Measurement sensors
cess, it is heated by the ground absorber and removes
humidity from the drying products. At night, part of the
In order to study the characteristics of the solar chimney
thermal energy stored by the ground during the solar radi-
as a solar dryer for agricultural products, the ambient con-
ation incidence period is transferred to the air?ow, allow-
ditions (temperature, humidity, wind velocity and solar
ing the continuous operation of the dryer.
radiation components) were measured and the thermal
As it is a new application for this device, the use of the
pro?le (velocity, humidity and temperature) of the air ?ow
solar chimney to dry agricultural products requires a techni-
inside the prototype unit was recorded.
cal feasibility study. This article aims to investigate the tech-
Six pyranometers (Eppley Black and White Model 8-48)
nical feasibility of the solar chimney in solar food drying.
were used to measure the di?use and global components of
Drying tests of co?ee grains, bananas, and tomatoes were
the incident solar radiation, inside and outside the device.
performed inside the dryer, and the results were compared
The uncertainty of the pyranometers was determined to
with natural sun drying, under the same climatic conditions.
be 5%, with a probability of 95%.
Two capacitive psychrometers were used to measure the
ambient relative humidity and the relative humidity of the
2. Experimental analysis
air?ow. The relative uncertainty of each psychrometer was
determined to be 6%, with a probability of 95%.
2.1. Prototype
The measurement of the ambient temperature, ground
temperature, and ?ow temperature was made using eight
A prototype solar chimney was built speci?cally for this
(8) k-thermocouples, with mineral insulation. The thermo-
study (Fig. 2). A tower of 12.3 m in height was constructed
couples were calibrated and an analysis resulted in an

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
Fig. 2. Solar chimney prototype.
uncertainty of measurement of the thermocouples of 1 °C,
allow the suitable statistic treatment of the results. Repeti-
with a probability of 95%.
tion tests were not performed due the random behavior of
For the measurement of the air?ow velocity, eight (8)
the solar energy and of the ambient conditions.
blade Homis anemometers, with a propeller of 50 mm in
diameter, were used. The uncertainty of measurement of
3. Results and discussion
the anemometers took into account the calibration and
the uncertainty of the air density correction (due to the
The experimental tests were performed from February
temperature and relative humidity). The global uncertainty
to November of 2003. The daily values of the extraterres-
of the anemometers was determined to be 6% with a prob-
trial solar radiation and of the global solar radiation mea-
ability of 95%.
sured (incident over the cover of the solar chimney) are
The analogical voltage or current signals from the
presented on Fig. 3. During the tests, the higher daily solar
humidity, temperature, velocity and solar radiation mea-
radiation measured on the cover was 28 ± 1 MJ/m2, occur-
surement sensors were converted into digital signals
ring in October (spring), while the lowest solar radiation
through ADAMS 4018 Modules in a data acquisition sys-
measured was 8.0 ± 0.4 MJ/m2, which occurred in the
tem. These modules have eight input analogical channels
same month.
(de?ning the maximum number of each kind of sensor),
Fig. 4 shows the values (daily maximum, minimum and
with an acquisition frequency of one sample per second.
average) of the ambient air temperature, for each day, com-
An analogical scale was used to assess product mass in
pared with the National Institute of Meteorology values
the performed drying tests. The absolute uncertainty of
for the ambient air temperature measured at Belo Horizon-
the plate scale was de?ned as the maximum calibration
te’s meteorological station (INMET, 2004). The di?erence
error: 0.3 g.
between the measured values of ambient air temperature
(compared to the values of the National Institute of Mete-
2.3. Drying tests
orology) can be explained by the distance between the mea-
surement locations and by the uncertainty of the
Drying tests of co?ee grains, whole bananas and toma-
instruments used. The yearly average ambient air tempera-
toes (cut into two parts) were performed. Before beginning
ture was 23 ± 1 °C. During the tests, the minimum ambient
the drying tests, each product received a pretreatment
air temperature (9 ± 1 °C) was observed in May (fall),
(according to Aguirre and Gasparino Filho, 1999). The
while the maximum ambient air temperature (41 ± 1 °C)
products were put into three immersion baths. The bana-
occurred in September (spring).
nas were hand peeled, the tomatoes were cut into two parts
The air?ow temperature was measured under the solar
and the seeds removed, and the pulp of the co?ee grains
chimney cover, at a radial position corresponding to r/
was removed. After these procedures, the products were
Rc = 0.15 (where Rc represents the cover radius) and at
divided into three samples. The ?rst sample was used to
an axial position corresponding to x/Hc = 0.50 (where
assess the initial moisture content (in a stove with forced
Hc represents the cover height). The daily minimum, aver-
circulation), the second was submitted to natural sun dry-
age and maximum air?ow temperatures are shown in
ing and the third sample was dried inside the solar chim-
Fig. 5. The maximum air?ow temperature (56 ± 1 °C)
ney. Each sample was subdivided into nine samples, to
occurred in February (summer), presenting a yearly

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
Fig. 3. Distribution of the daily solar radiation.
Fig. 4. Daily minimum, average and maximum ambient air temperature for the performed tests.
maximum ?ow temperature average of 42 ± 1 °C. It is
air?ow allowed the estimation of the drying capacity of
important to note that the maximum average temperatures
the physic model. According to Leon et al. (2002), the ideal
were obtained when the average incident solar radiation
drying capacity of a solar dryer is 4 kg per 0.0125 m3/s of
had reached its maximum values. During the year, it was
air?ow. Considering the yearly average mass ?ow of
observed a maximum increase of the air?ow temperature
1.40 ± 0.08 kg/s, the drying capacity of the built physic
over the ambient air temperature of 27 ± 1 °C, reached in
model is approximately 440 kg.
A yearly average ambient air relative humidity of
Fig. 6 shows the maximum, minimum and average val-
63 ± 4% was obtained. The minimum ambient air relative
ues of mass ?ow, during the testing periods. The maximum
humidity (21 ± 1%) occurred in March, whereas the maxi-
mass ?ow (2.8 ± 0.2 kg/s) was observed in November
mum ambient air relative humidity (91 ± 5%) occurred in
(spring), while the minimum mass ?ow (0.70 ± 0.04 kg/s)
September. The yearly average air?ow relative humidity
was observed in July (winter). The yearly average of the
measured was 54 ± 3%. The minimum air?ow relative
mass ?ow was 1.40 ± 0.08 kg/s. The mass ?ow of the hot
humidity was 8.0 ± 0.5% (in March), while the maximum

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
Fig. 5. Daily minimum, average and maximum ?ow temperature for the performed tests.
Fig. 6. Minimum, average and maximum daily mass ?ow for the performed tests.
value was 87 ± 5% (in October). When the yearly average
of the device, higher air?ow velocities were also found near
for the air?ow’s relative humidity in the prototype
the tower. These higher values of velocity and temperature
(54 ± 3%) is compared to the yearly average for the ambi-
in the central area improved the drying the process, indicat-
ent air’s relative humidity (63 ± 4%), the advantage of the
ing that the central area is the most suitable place to posi-
drying on the solar chimney becomes clear. This occurs due
tion the products to be dried.
to the increase in temperature and the reduction of the
Drying tests of co?ee grains, bananas, and tomatoes
?ow’s relative humidity, thus reducing the equilibrium
were performed. In all the tests, the time required for dry-
moisture content and increasing the free humidity and
ing inside the solar chimney was lower than that required
the drying velocity (Aguirre and Gasparino Filho, 1999).
for natural sun drying. It is important to note that the
The temperature pro?le measured under the cover indi-
desired ?nal moisture content was reached for the products
cates that the higher temperature occurred at the center,
dried in the solar chimney.
near the tower and the ground surface. As the ?ow area
The drying curve of the co?ee grains is shown in Fig. 7.
was decreased in the radial direction towards the center
The co?ee grains needed 152 h to be dried when directly

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
Fig. 7. Drying curve of co?ee bean dried using the solar chimney and natural sun drying.
exposed to the sun, while the time required in the solar
tant to note that the solar chimney presented a greater
chimney was only 76 h, approximately half of the time
reduction in drying time. The time required for natural
required for natural sun drying. For the co?ee grains, the
sun drying of the products to desired moisture content
initial moisture content in wet basis was 50 ± 3%, whereas
was reduced by 72%, in contrast to the 76% reduction
the ?nal moisture content was 11.0 ± 0.7%. During the
required for the tunnel solar dryer created by Schirmer
tests performed, the solar energy incident on the cover
et al. (1996). Purohit et al. (2006) developed a framework
was 63 ± 4 MJ/m2, the average mass ?ow was 1.28 ±
to facilitate the comparison of the ?nancial feasibility of
0.08 kg/s, the average air?ow temperature was 21 ± 1°C
the solar drying of any agricultural product for potential
and the average ambient air temperature was 32 ± 1°C.
users of natural sun drying. The study performed by Puro-
Mwithiga and Kigo (2006) built a small solar dryer, used
hit et al. (2006) showed that the comparison of the drying
to dry co?ee grains. When compared to natural sun drying,
times (solar chimney vs. natural sun drying) represents a
the drier reduced the time required to dry the co?ee grains
suitable methodology to assess solar driers.
by 60%, reaching a ?nal moisture content of 13% (w. b.).
Fig. 9 shows the tomato’s drying curve in the sun and in
The solar chimney allowed a reduction of 50%. Neverthe-
the solar dryer. The natural sun drying occurred over a per-
less, the solar chimney allows a greater drying capacity,
iod of 195 h, while the drying of the tomatoes inside the
rendering it more suitable to drying the amount of grains
solar chimney occurred in 67% of this time (130 h). The
produced in Brazil.
tomato’s drying process occurred under the same condi-
Fig. 8 compares the banana’s drying curves inside the
tions of the banana’s drying process.
solar dryer with natural sun drying (to a ?nal moisture con-
Microbiological contamination of the dried products
tent of 25% in a wet basis). The natural sun drying was
was not observed. All dehydrated products presented
completed in 193 h, while the drying inside the chimney
acceptable ?avor, texture and color.
was completed in 139 h (72% of the time spent for the nat-
Despite obtaining a shorter drying time using the solar
ural sun drying). During the tests the solar energy incident
chimney, a global analysis of the energy absorbed by the
on the cover was 84 ± 5 MJ/m2, the average mass ?ow was
air?ow, as compared to the solar incident energy, resulted
1.36 ± 0.08 kg/s, the average air?ow temperature was 20 ±
in a low e?ciency for the device (approximately 7%). This
1°C and the average ambient air temperature was
low e?ciency can be attributed to signi?cant heat losses,
30 ± 1°C. Schirmer et al. (1996) dried slices of bananas
occurring mainly due to the inner layers of the ground,
(1 cm thick) in a tunnel solar drier and compared the
as indicated by local energy balances. It can be observed
results with natural sun drying. To achieve a ?nal moisture
that a considerable amount of the absorbed solar radiation
content of 30% (w.b.), the banana slices required 84 h in
and of the stored energy in the ground does not return to
the tunnel solar drier and 110.5 h exposed directly to the
the ?ow at night. The use of a plastic material at the cover
sun (for a daily incident solar energy of 23.8 MJ/m2 day).
also contributes to signi?cant energy losses. The plastic
Despite di?erent drying conditions (bananas shape, ?nal
used in the cover presents a low transmittance for solar
moisture content, and incident solar radiation), it is impor-
radiation (72 ± 5%) and a high transmittance for infrared

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
Fig. 8. Drying curve of bananas dried using the solar chimney and natural sun drying.
Fig. 9. Drying curve of tomatoes dried using the solar chimney and natural sun drying.
radiation emitted by the warmed ground (40.0 ± 1.5%).
and thermal ?uid dynamics of the hot air ?ow generated
Moreover, the device e?ciency can be sensitively increased
was performed. The solar chimney is a device technically
using a layer of thermal insulation under the ground and
feasible for the drying of food, mainly grains. The perfor-
replacing the cover material with one with higher transmit-
mance of the solar chimney can be modi?ed by equipping
tance for solar radiation and a lower transmittance for
the device with the proposed improvements, in accordance
infrared radiation emitted by the warmed ground. A suit-
with each farmer’s speci?c needs.
able material should be glass.
The tower, made of wood and ?berglass, proved to pro-
vide a good thermal insulation. A visual assessment of its
4. Concluding remarks
physical integrity suggests a great durability for this com-
ponent. Due to the higher cost required for the tower con-
This paper presents an experimental study of a solar
struction, the replacement of the constructive materials
chimney. To assess its use for food drying, a prototype of
should be studied. The thermal di?user plastic ?lm presents
a solar chimney was built in Belo Horizonte (Brazil). An
low weight and reduced costs, and is more suitable to be
experimental simulation of the ambient thermal conditions
used in the solar chimney’s cover.

A.G. Ferreira et al. / Solar Energy 82 (2008) 198–205
The thermal behavior of the ?ow shows that the drying
Chen, H.-H., Hernandez, C.E., Huang, T.-C., 2005. A study of the drying
products must be placed at the center of the device, close to
e?ect on lemon slices using a closed-type solar dryer. Solar Energy 78,
the ground. Drying tests on bananas support these veri?ca-
Condor?´, M., Saravia, L., 2003. Analytical model for the performance of
tions and suggest that drying on wire mesh is more interest-
the tunnel-type greenhouse drier. Renewable Energy 28, 467–485.
ing than drying on plastic canvas; however, the use of
El-Sebaii, A.A., Aboul-Enein, S., Ramadan, M.R.I., El-Gohary, H.G.,
superposed meshes actually delays the drying process.
2002. Experimental investigation of an indirect type natural convection
The constructed prototype showed a maximum temper-
solar dryer. Energy Conversion and Management 43, 2251–2266.
Forson, F.K, Nazha, M.A.A., Rajakaruna, H., 2007. Modelling and
ature increase of 27 ± 2 °C (related to the ambient air tem-
experimental studies on a mixed-mode natural convection solar crop-
perature) and a reduction of the ?ow’s relative humidity (in
dryer. Solar Energy 81, 346–357.
terms of the yearly average) from 63 ± 4% to 53 ± 4%.
Hossain, M.A., Bala, B.K., 2007. Drying of hot chilli using solar tunnel
This ?ow temperature increase and the reduction of the
drier. Solar Energy 81, 85–92.
?ow’s relative humidity (related to the ambient relative
INMET – Instituto Nacional de Meteorologia, 2004, available on
<>, in 02 january, 2004.
humidity) ensure the feasibility of the device as a solar
Ivanova, D., Enimanev, K., Andonov, K., 2003. Energy and economic
dryer, presenting advantages when compared to natural
e?ectiveness of a fruit and vegetable dryer. Energy Conversion and
sun drying. During the tests performed in 2003, an average
Management 44, 763–769.
yearly mass ?ow of 1.40 ± 0.08 kg/s could be observed,
Leon, M.A., Kumar, S., Bhattacharya, S.C., 2002. A comprehensive
allowing one to estimate the drying capacity of the device
procedure for performance evaluation of solar food dryers. Renewable
and Sustainable Energy Reviews 6, 367–393.
to be 440 kg of product (320 kg of products per m3/s of
Madhlopa, A., Ngwalo, G., 2007. Solar dryer with thermal storage and
air ?ow).
biomass-backup heater. Solar Energy 81, 449–462.
The low thermal e?ciency observed can be explained
Mu¨hlbauer, W., Mu¨ller, J., Esper, A., Bux, M., 1996. Secagem Solar e ao
mainly by the heat di?usion through the ground, by the
Sol para Produtos Agr?´colas e Florestais, Universidade de Hohen-
low transmittance of solar radiation and by the high trans-
heim/Instituto para Engenharia Agr?´cola nos Pa?´ses Tropicais e
Subtropicais, Stuttgart/Alemanha.
mittance of infrared radiation from the plastic cover. These
Mwithiga, G., Kigo, S.N., 2006. Performance of a solar dryer with limited
losses can be minimized by implementing thermal insula-
sun tracking capability. Journal of Food Engineering 74 (2), 247–252.
tion in the ground and replacing the plastic material of
Pangavhane, D.R., Sawhney, R.L., Sarsavadia, P.N., 2002. Design,
the cover.
development and performance testing of a new natural convection
solar dryer. Energy 27, 579–590.
Purohit, P., Kumar, A., Kandpal, C., 2006. Solar drying vs. open sun
drying: a framework for ?nancial evaluation. Solar Energy 80, 1568–
Aguirre, J.M., Gasparino Filho, J., 1999. Instituto de Tecnologia de
Schirmer, P., Janjai, S., Esper, A., Smitabhindu, R., Mu¨hlbauer, W., 1996.
Alimentos. Desidratac¸a˜o de frutas e hortalic¸as: Manual te´cnico.
Experimental investigation of the performance of the solar tunnel
dryer for drying bananas. Renewable Energy 7 (2), 119–129.
Bena, B., Fuller, R.J., 2002. Natural convection solar dryer with biomass
Simate, I.N., 2003. Optimization of mixed-mode and indirect-mode
back-up heater. Solar Energy 72 (1), 75–83.
natural convection solar dryers. Renewable Energy 28, 435–453.

Document Outline
  • Technical feasibility assessment of a solar chimney for food drying
    • Introduction
    • Experimental analysis
      • Prototype
      • Measurement sensors
      • Drying tests
    • Results and discussion
    • Concluding remarks
    • References