ARTICLE IN PRESS

BIOSYSTEMS ENGINEERING 98 (2007) 33– 38

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/issn/15375110

Research Paper: PH—Postharvest Technology

Development of a dryer with airflow reversal and a

pneumatic system for grain movement

C.P. Sampaioa,, R.M. Nogueirab, C.D. Robertob, J.S. Silvab

a
CNPq/SEPLAN/MCT/ULBRA, Palmas, TO 77054-970, Brazil
b
Department of Agricultural Engineering, Universidade Federal de Vic- osa, Vic- osa, MG 36571-000, Brazil

art i cle info
Article history:
Received 23 February 2006
Accepted 21 February 2007
Available online 2 July 2007
The objectives of this work were elaboration, construction and evaluation of a new dryer
model for coffee. For evaluation, three tests were conducted for each type of coffee. In tests
with natural coffee, the drying times were 11.5, 10.3 and 9.6 h. The coffee with initial
moisture content of 24.671.8% w.b. was dried to storage moisture of 11.971.3% w.b. For the
parchment coffee, drying times were 14.0, 12.5 and 12.2 h. The coffee with initial moisture
content of 33.972.1% w.b. was dried to storage moisture of 12.071.5% w.b. The drying time
of natural coffee was less because the moisture content of natural coffee was less in
comparison with that of parchment coffee. The coffee dried with the new drying system
was considered as a coffee of high cup quality, and was classified as types 6 and 4 for
natural and parchment coffee, respectively.
& 2007 Published by Elsevier Ltd. on behalf of IAgrE

1. Introduction The mechanical drying consists of submitting the product

As described by Silva et al. (2000), drying coffee is, compara-
tively, difficult to execute, because of the high initial moisture
content, which is usually around 60% w.b., and the sugar-rich
mucilage. Therefore, the possibility of fermentation during
the first drying stage is high, and this may affect adversely the
product quality.
The physical characteristics of coffee fruits or coffee berries
are important for developing equipment. Information about
volume, superficial area, shape and dimensions of a product
are important for studies on mass and heat transfer as well as
airflow within a granular mass of the product (Afonso Jr et al.,
2003). Some problems in drying coffee occur because the
dryers are designed for other grains and adapted for coffee,
resulting in high operational cost and low energy efficiency
(Pinto Filho, 1994).
to the action of an airflow that passes through grains mass.
According to Silva et al. (2001), use of mechanical dryers for
coffee beans presents several advantages in relation to the
solar drying on a cement terrace. Designing and constructing
of efficient drying systems that deliver good final quality
product should be seen as technological development as it
reduces the drying costs and to increase the competitiveness
of the Brazilian coffee in the international market. For
Brooker et al. (1992), grains drying systems, used since 1950,
are more evolutionary than revolutionary. In Brazil, research-
ers and industry are applying known technologies to adapt or
develop drying systems that fit the investment capacity and
work conditions of Brazilin coffee growers (Donzeles, 2002).
The fixed-bed dryer, a model frequently used for the pre-
drying or drying of coffee, requires stirring for product
homogenisation during drying. When manually executed,

Corresponding author.
E-mail addresses: cpsampaio@ulbra-to.br (C.P. Sampaio), juarez@ufv.br (J.S. Silva).
1537-5110/$ - see front matter & 2007 Published by Elsevier Ltd. on behalf of IAgrE
doi:10.1016/j.biosystemseng.2007.02.014
 ARTICLE IN PRESS
34 BIOSYSTEMS ENGINEERING 98 (2007) 33 – 38

the coffee stirring operation needs great physical effort,
mainly at the beginning of the drying process, when the
product has high moisture (De Grandi, 1999).
For the dryers with mechanical stirring, increasing bean
flow in the drying chamber gives, according to Dalpasquale et
al. (1991), in general, a better final product quality. However,
increasing the coffee bean velocity increases the specific
energy consumption and decreases drying efficiency, because
coffee beans at a higher speed in the drying chamber achieve
less decrease of moisture content per unit of time.
The pneumatic conveying is a simple transport system with
relatively low cost. The conveyors use high-speed air to move
the grains by a hermetic piping system, can transport the
product in any direction, including a curved pathway (Silva,
2000).
Therefore, by combining the advantages of the fixed-bed
dryer with the simplicity of the pneumatic conveyor system,
the objective of this work is to develop and evaluate of a
coffee drying system that allows for product movement
(loading, unloading and stirring) using only one fan, respon-
sible for grain movement and the grain drying airflow.
2. Material and methods
ber. The controls for drying, stirring and unloading of the
dryer were manually operated by the registers located
along the piping. The conveying system, due to available
facilities, was built of square section, with a length of
6.7 m, transverse sectional area of 0.01 m2 and volume of
0.067 m3. The main characteristics of the pneumatic
system is shown in Table 1.
(3) Feed hopper
Besides receiving the flow of beans from the pre-drying
stage, the feed hopper was used to distribute the grain in
the drying chamber and was provided with a rubber pipe
shock absorber for reduction of impact grains. Its volume
was 0.30 m3.
(4) Drying chamber
The drying chamber was a special structure composed of
two pyramids constructed with perforated foils (23%
perforation) and connected at the bases. The structure
allowed the airflow, in direct contact with the coffee
beans, to dry in a reverse flow. With basal area of 1.10 m2
and height of 0.48 m, the pyramids can hold 0.12 m3 of
coffee beans. To maintain evenly distributed coffee in the
drying chamber, two special devices were built. With
pyramidal forms, the mixing devices were put on the
external part of the upper pyramid and, internally, in the

The coffee drying system was built of metallic foils and

masonry bricks, shown in Fig. 1, and consists of.
Table 1 – Main characteristics of the pneumatic system

(1) Charge hopper

Characteristics Air Air+product
A cylindrical hopper with height 1.09 m and diameter
0.6 m contains coffee beans to charge the dryer during Total pressure, mm[H2O] 71.2 72.4
drying process. The wall of the conical base was inclined Static pressure, mm[H2O] 32.0 40.0
at 351, giving a lower section a height of 0.13 m and a Air velocity, m s 1 25.3 23.0
diameter of 0.6 m, whose volume is 0.32 m3. Airflow rate, m3 min 1 15.2 13.8
Airflow density, m3 min 1 m 2
1.52 1.38
(2) Conveying system
Product velocity, m h 1 — 21.6
The conveying system comprises of a metallic piping for
Transport capacity, kg h 1 — 318.92
grain stirring and airflow conveying to the drying cham- Fan power, kw 0.38 —

Feed
hopper

Charge
hopper
Drying Unloading
chamber spout

Furnace Plenum

Fig. 1 – Coffee dryer with pneumatic conveying system.

 ARTICLE IN PRESS
BIOSYSTEMS ENGINEERING 98 (2 007 ) 3 3 – 38 35

lower pyramid. The devices for grain mixing were
composed of parallel bars (Fig. 2). In addition to allowing
for good distribution of the product in the pyramids, it
established repose angles. This allowed for gravity to
move the product and it occupied the external areas of the
upper pyramid and the internal areas of the lower. With
the coffee beans evenly distributed, the drying chamber
took the form of a hollow octahedron, with walls of 5.0 cm
thickness, forming two plenum chambers.
(5) Dryer unloading
The dryer is unloaded through an unloading valve, by
blocking the airflow from the plenum chamber and also the
flow of beans from the feed hopper. Worked manually, the
transportation system can be used as a stirring device,
during drying or as an unloading device once drying is
completed.
As coffee beans have high moisture at harvest, pre-drying
of the natural and parchment coffee was done in a terrace
dryer as recommended by Silva et al. (2003).
For the purposes of drying tests, initial moisture content of
coffee beans was 24.371.6% w.b. for the natural coffee and
33.972.4% w.b. for the parchment coffee. The drying was
done with the complete batch in the drying chamber. The
total coffee mass was divided in two batches (1 and 2), so one
batch stayed in the drying chamber and the other in the feed
hopper. In other words, while one batch is maintained in
repose, the other is in the drying stage. Alternating drying and
repose, the product is maintained in the system until reach-
ing a moisture content of 12% w.b., is reached characterising
the process as intermittent. When moisture content for
storage was reached, the two batches (1 and 2) were mixed
in continuous flow and, they remained in the dryer until they
reached the moisture content of 12% (w.b.).
The temperature of the air for drying was around 6072.8 1C.
This temperature of air is important for the temperature of
the coffee beans mass to exceed 40 1C, for otherwise the ‘cup
quality’ is affected. The product was stirred inside of the dryer
each hour during the drying period. Data on the product final
moisture content were taken after the coffee stirring, by the
means of the oven method.
Evaluation of dryer performance was based on methodol-
ogy proposed by Bakker-Arkema et al. (1978). However, the
most important parameters measured following Osório (1982)
were: temperature of the grain mass, relative humidity and
temperature of the ambient air and exhaust air of the dryer,
initial and final moisture content of the coffee beans, flow
rate of drying air, drying time, heat capacity of the fuel, and
product velocity inside dryer.

3. Results and discussions

In Table 2, values of the coffee beans mass are presented in

the initial and final phases for each test, as well as the bulk
density values (tw) determined in a test weight scale (1.0 l
capacity). In this table, it is observed that the coffee masses
varied in agreement with the reduction of the moisture
content and that the drying time varied for natural and

Fig. 2 – Details of the drying chamber.

Table 2 – Beans properties for natural and parchment coffees before and after drying

3
Test no Beans mass, Kg Bulk density, kg m Moisture, % (w.b.) Drying time, h

Initial Final Initial Final Initial Final

Natural coffee
1 136 116 595.6 452.2 25.8 11.8 11.5
2 156 132 594.3 451.3 25.0 11.9 10.3
3 156 136 596.0 452.8 23.1 11.9 9.6

Parchment coffee
4 232 166 476.0 440.4 36.8 11.9 14.0
5 148 116 452.3 438.5 32.5 12.0 12.5
6 182 142 451.9 437.1 32.3 12.0 12.3
 ARTICLE IN PRESS
36 BIOSYSTEMS ENGINEERING 98 (2007) 33 – 38

parchment coffees. The drying times of the natural coffee for
tests 1, 2 and 3 were 11.5, 10.3 and 9.6 h, respectively. The
reduction in coffee moisture content for theses tests were
14% (from 25.8% to 11.8% w.b.) for test 1, 13.1% (from 25.0%
to 11.9% w.b.) for test 2 and 11.2% (from 23.1% to 11.9% w.b.)
for test 3.
For the parchment coffee (tests 4, 5 and 6), drying time was
14.0 h for test 4, 12.5 h for test 5 and 12.3 h test 6. Initial
moisture contents were 36.8%, 32.5% and 32.4% w.b., and the
final moisture contents were 11.9%, 12.0% and 12.0% w.b.,
respectively.
28
Moisture Content, % (w.b.)
A longer drying time (14.0 h) was observed in test 4 for the
parchment coffee because of the larger initial mass (232 kg)
and higher initial moisture content of this batch compared to
other tests. Test 3, with natural coffee, presented a smaller
drying time (9.6 h), which is linked to the psychometric
properties of the ambient air moving with the drying air
and with the lowest initial moisture content (23.1% w.b.). In
any case, drying of parchment coffee was shown to be
advantageous in all of the tests, because initial parchment
coffee mass almost double of the initial natural coffee mass.
Still, they are noted as higher moisture contents for parch-
ment coffee, with small increases in drying time.
The curves presented in the Figs. 3 and 4 show the
conditions of the coffee beans and the corresponding drying
times. In these curves, coffee drying period is represented by

26 the full black lines. The holding period (without contact with
24 drying air), found among the drying intervals, is represented
22 by broken lines. It is also observed that the final portion of the
20 curve (dotted line) shows that the drying is accomplished
18 with the continuous movement of the coffee beans, for mass
16 homogenisation, knowing that this achieved moisture con-
14 tent is recommended for storage. As observed by Cordeiro
12 (1982), during each holding or rest period, inside the feed
10 hopper, a small reduction in the moisture content of the
0 1 2 3 4 5 6 7 8 9 10 11
coffee mass, of 0.5%, on average, was noted.
Time, h
In the Fig. 3, a larger reduction of the moisture content of
Fig. 3 – Moisture reduction of natural coffee drying: ——, the coffee mass was observed in the first hour of drying, since
drying; , holding; - - - -, homogenization. moisture content and initial mass were larger, compared to
other tests with natural coffee. Throughout the process, that
loss was less accentuated. Once conditions of lower moisture
40 content are reached, there is more resistance to the loss of the
Moisture Content ,% (w.b.)

water.
35
In Table 3, it can be verified that the specific energy
30 consumption for the tests 1, 2 and 3, using natural coffee,
25 were 38.2, 26.4 and 36.0 MJ kg 1, respectively. In tests 4, 5 and 6
(parchment coffee) the specific consumptions were 13.1, 26.5
20 and 15.7 MJ kg 1, respectively.
15 The smallest specific energy consumption of fuel in test 4
was the result of higher initial moisture content of product
10
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 (36.8% w.b.) which permitted a higher drying rate at the
beginning of the process. It is noted, however, that the drying
Time, h
time was longer for this test.
Fig. 4 – Moisture reduction of parchment coffee drying: ——,
drying; , holding; - - - -, homogenization.

Table 3 – Drying time, initial mass and energy consumption

1
Test no Drying time, h Water loss, kg Energy consumption, MJ Specific energy consumption, MJ kg

Fuel Electric Fuel Electric Total

Natural coffee
1 11.5 20.0 701.7 61.5 35.08 3.1 38.2
2 10.3 24.0 585.9 48.9 24.4 2.0 26.4
3 9.6 20.0 663.6 54.5 33.2 2.7 36.0

Parchment coffee
4 14.0 66.0 797.6 72.9 12.1 1.1 13.1
5 12.5 32.0 799.8 68.0 24.4 2.1 26.5
6 12.2 46.0 671.1 55.3 14.5 1.2 15.7
 ARTICLE IN PRESS
BIOSYSTEMS ENGINEERING 98 (2 007 ) 3 3 – 38 37

40

35

Energy consumption, MJ kg−1

30

25

20

15

10

0
Test 1 Test 2 Test 3 Test 4 Test 5 Test 6
Fig. 5 – Comparison of the specific energy consumption electric and fuel (coal), for the stirring of the product with warm air
(&) and with ambient air ( ).

Table 4 – Results of cup quality and other coffee

characteristics
Test no Cup quality Type
Natural coffee
1 Hard/Green 6 Regular
2 Hard/Green 6 Regular
3 Hard Green 6/7
Parchment coffee
4 Soft only 4 Regular
5 Soft only 4 Regular
6 Soft only 3/4
As is evident, the prototype presented high energy con-
sumption. This can be explained by the product stirring
using the heated drying air. However, when stirring with
ambient air, an energy economy of 24.1% was verified
for natural coffee and 31.9% for parchment coffee, in relation
to the total energy consumption in Table 3. The difference
in the energy consumption can be observed graphically in
Fig. 5.
Because the airflow used to dry the product was the same
as that used in the pneumatic transportation process, the
amount of drying air was much higher than the optimum
airflow. As stated by Nalladurai et al. (2002), to design drying
and aeration systems, data on airflow resistance of grains is
necessary and fans also must be selected. For good energy
efficiency, during drying (with the proposed dryer) the airflow
used in the transport phase must be partially blocked to avoid
large heat losses in exhaust air.
The samples of natural and parchment coffee were sent to
COOXUPÉ—Regional Cooperative of the coffee growers
in Guaxupé-MG., for coffee beans classification and sen-
sorial analyses or ‘cup quality’. The results are provided in
Table 4.
4. Conclusions
Flavour Appearance
Based on the results of the dryer prototype evaluation, coffee
classification and cup quality of the product, it is concluded
Good
that:
Good
Regular Good
(1) there was no adverse effect on product quality from any
Good mechanical damage sustained from impacts caused by
Good the pneumatic system;
Regular Good (2) the temperature of the coffee beans mass was lower than
the drying air temperatures, with an average value of
36.2571.6 1C;
(3) the drying time varied significantly between treatments of
natural and parchment coffees;
(4) the tests with coffee beans with higher initial moisture
contents present larger efficiencies in the drying process;
(5) the coffee holding period in the feed hopper within the
drying process is very important for acheiving a decrease
in the coffee moisture content, on average, 0.5%;
(6) the prototype presents high average specific energy
consumption, water removal measures were 33.5 MJ kg 1
and 18.4 MJ kg 1 of water removed for natural coffee and
for parchment coffee, respectively;
(7) the high specific energy consumption and the low energy
efficiency can be accounted for by the high warm airflow
in the tests, which caused the air in the exit system to
have high drying potential; and
(8) the coffee dried with the new drying system was con-
sidered of high ‘cup quality’ and was classified as types 6
and 4 for natural and parchment coffee, respectively.
R E F E R E N C E S
Afonso P C; Corrêa P C; Pinto F A C; Sampaio C P (2003). Shrinkage
evaluation of five different varieties of coffee berries during
the drying process. Biosystems Engineering, 86(4), 481–485
 ARTICLE IN PRESS
38 BIOSYSTEMS ENGINEERING
Bakker-Arkema F W; Lerew L E; Brook R C; Brooker D B (1978).
Energy and capacity performance evaluation of grain dryers.
ASAE Paper No. 78-3523, St Joseph, MI
Brooker D B; Bakker-Arkema F W; Hall C W (1992). Drying and
storage of grains and oilseeds. An AVI book, New York, NY
Cordeiro J A B (1982). Influência da temperatura e do tempo de
repouso na secagem de café (Coffea arabica L.) em camadas
fixas. [Influence of temperature and repose time in coffee fixed
bed drying.] Master Thesis, Universidade Federal de Vic-osa,
Vic- osa, Brazil
Dalpasquale V A; Pereira D A M; Sinicio R; Oliveira Filho D (1991).
Secado de granos a altas temperaturas. [High temperature
grain drying.] Oficina Regional de la Fao para America Latina y
el Caribe, Santiago, Chile
De Grandi A M (1999). Avaliac-ão da eficiência de secagem de café
(Coffea arábica L.),em secador de camada fixa vertical com
revolvimento mecânico. [Evaluation of drying efficiency for a
vertical fixed bed coffee dryer with mechanical stirring]
Master Thesis, Universidade Federal de Vic-osa, Vic-osa, Brazil
Donzeles S M L (2002). Desenvolvimento e Avaliac-ão de um
sistema hı́brido, Solar e biomassa, para secagem de café
(Coffea arabica L.). [Development and Evaluation of a hybrid
system, solar and biomass, for coffee drying (Coffea arabica L.)]
Doctorate Thesys, Universidade Federal de Vic-osa, Vic-osa,
Brazil
98 (2007) 33 – 38
Nalladurai K; Alugusundaram K; Gayathri P (2002). Airflow
resistance of paddy and its byproducts. Biosystem Engineer-
ing, 83(1), 67–75
Osório A G S (1982). Projeto e construc-ão de um secador
intermitente de fluxos concorrentes e sua avaliac-ão na
secagem de café. [Desing and construction of a intermittent
concurrent flow dryer and its evaluation for coffee drying.]
Master Thesis, Universidade Federal de Vic-osa, Vic-osa, Brazil
Pinto Filho G L (1993). Desenvolvimento de um secador de fluxos
cruzados com reversão do fluxo de ar de resfriamento, para a
secagem de café (Coffea arabica L.). [Development of a crossed
flow dryer with reversion of cooling air flow for coffee drying.]
Master Thesis, Universidade Federal de Vic-osa, Vic-osa, Brazil
Silva J S (2000). Secagem e armazenagem de produtos agrı́colas.
[Drying and storage of agricultural products.] Editora Aprenda
Fácil, Vic- osa, Brazil
Silva J S; Sampaio C P; Machado M C; Lo Monaco P A (2000).
Preparo, Secagem e Armazenagem. [Preparation, drying and
storage.] In: (Silva J S, ed), Secagem e armazenagem de café-
tecnologias e custos. [Coffe drying and storage—tecnologies
and costs.] Vic-osa, Brazil
Silva J S; Nogueira R M; Lopes R P (2003). Um Sistema ideal para
secagem e armazenagem de café. [A ideal system for coffee
drying and storage.] In: (Zambolim L, ed), Produc- ão integrada
de café. [Integrated production of coffee.] Vic- osa, Brazil