The use of biomass residues in the Brazilian soluble coffee industry

Abstract

The objective of this paper is to discuss the use of coffee grounds in the Brazilian soluble coffee industry. This residue is used as a fuel in the boilers of the same industry; so, data about their utilization are presented and analysed, discussing the actual technology and the advantages of improving the drying of the biomass with the exhaust combustion gases. After that, an experimental study is reported on the characteristics of this material, which are important for the combustion process, including the transport, storage and drying, the mean diameter of the particles, talus angle, apparent and real density, sphericity, surface area, terminal velocity, spontaneous ignition temperature and heat of combustion.

Full text

THE USE OF BIOMASS RESIDUES IN THE BRAZILIAN
SOLUBLE COFFEE INDUSTRY
M.A. SILVA*},S.A. NEBRA$,M.J. MACHADO SILVA$and C.G. SANCHEZ%
*DTFD, FEQ, UNICAMP, P.O. Box 6122, 13083-970, Campinas, S.P., Brazil
$DE, FEM, UNICAMP, P.O. Box 6122, 13083-970, Campinas, S.P., Brazil
%DETF, FEM, UNICAMP, P.O. Box 6122, 13083-970, Campinas, S.P., Brazil
(Received 3 January 1997; revised 17 October 1997; accepted 21 October 1997)
AbstractÐThe objective of this paper is to discuss the use of coee grounds in the Brazilian soluble cof-
fee industry. This residue is used as a fuel in the boilers of the same industry; so, data about their utiliz-
ation are presented and analysed, discussing the actual technology and the advantages of improving the
drying of the biomass with the exhaust combustion gases. After that, an experimental study is reported
on the characteristics of this material, which are important for the combustion process, including the
transport, storage and drying, the mean diameter of the particles, talus angle, apparent and real density,
sphericity, surface area, terminal velocity, spontaneous ignition temperature and heat of combustion. #
1998 Published by Elsevier Science Ltd. All rights reserved
KeywordsÐResidues; burning; coee grounds; terminal velocity; spontaneous ignition; combustion
heat.
1. INTRODUCTION
The necessity of conserving nature and the
increasing consumption of energy resources
have led researchers to the study of renewable
sources of energy. Among these renewable
sources, the combustion of the following agri-
cultural residues are some of the most import-
ant: ears of corn, rice husks, cotton twigs,
coconut ®bres, jute stems and wheat straw.
1,2
There are also agricultural industrial residues
such as sugar cane bagasse
1
and coee
grounds (Dutra, personal communication).
3,4
All of these residues can have other uses, and
opting for one utilization over another
depends on a global analysis which takes into
account the economic, social and environmen-
tal aspects.
The coee grounds, which are the subject of
this paper, are the residues of the soluble cof-
fee industry. Soluble coee, without the ad-
dition of carbohydrates, was ®rst produced in
about 1950 in the U.S.A.,
5
and its consump-
tion has increased since then, reaching 19% of
the total coee consumed in the world in
1980.
6
As a consequence, the production of
coee grounds has increased, and their dispo-
sal has become a matter of increasing con-
cern.
Brazil began to produce soluble coee in
1962, and it has been a major world producer
since 1970, when production reached 36
million kilograms per year.
7
In 1989, there
were 11 plants (ABICS, personal communi-
cation), with 85% of the production going to
the external market. In 1988, only 11.8% of
the coee exported was in the form of soluble
coee, which totalled 120 million kilograms
with a value of 213 million dollars.
5
Initially, 1.86 kg of spent coee grounds
were produced for each kilogram of soluble
coee.
5
With improvements to the industry,
this ®gure dropped to 1.27 kg in 1967,
7
and it
is now 0.91 kg.
8
The daily volume of coee
grounds production is therefore high, consid-
ering that after the extraction of the soluble
coee, it has a humidity of 75±80% (w.b.),
which is reduced to 50% after pressing
(Dutra, personal communication).
1,3,5,9
Coee grounds are highly pollutant due to
the presence of organic material that demands
a great quantity of oxygen in order to
degrade. Simply piled up, they can ferment
and produce spontaneous combustion, as has
occurred in some storage sites (Dutra, per-
Biomass and Bioenergy Vol. 14, Nos. 5/6, pp. 457±467, 1998
#1998 Published by Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0961-9534/98 $19.00 + 0.00
PII: S0961-9534(97)10034-4
}Corresponding author.
457

sonal communication). They can thus not be
thrown away untreated.
3,10
With such a pollutant industrial residue
being produced in great quantity, the identi®-
cation of more rational uses has become
necessary. To study the feasibility of these
uses it is necessary to know the composition
of the coee grounds, which is reported in
Table 1. The reported composition varies from
one author to another, such discrepancies
being due to the dierent varieties of coee
used in production.
The ®rst attempt was to use coee grounds
as a fertiliser,
4,11,12
however, due to its low
nitrogen content (approximately 2%) and its
high acidity (approximate pH of 4.2),
10
this
use was considered uneconomic.
13
On analysis of the amino acids contained in
the protein of the coee grounds, it was found
that half of the essential amino acids are
absent, so a complement would be necessary
in order for the coee grounds to be used as
animal feed.
4
For this reason the idea was
abandoned.
3,10,12
Others uses were researched: its utilization
as a substitute for wood powder, the pro-
duction of methanol and acetone,
3
the extrac-
tion of oil;
4,12
to make microporous materials,
packing materials; exchangers of ions,
4
in the
production of cigarette ®lters, resins, special
lubricators, as a polisher for painting and a
carrier for insecticides and herbicides.
10
All these other possibilities have been shown
to be infeasible or to require long and expens-
ive programmes of development. All the
authors are in agreement on this aspect, con-
®rming its use as a fuel as the best utilization
of this residue. Burning also permits the ad-
dition of the centrifugation residue of the
liquid current from the press, which has 2±7%
of solids and can not be disposed of in rivers,
because of its high demand for oxygen, as a
result of which it is highly pollutant.
Moreover, the ashes from burning are good
fertilisers, because of their composition which
has a high content of phosphorus, calcium
and magnesium.
4,5
The heat value of coee grounds is similar
to coal and is higher than wood and other
biomass, as can be seen in Table 2, in which a
comparison is made.
The usual process is as follows: when the
spent coee grounds leave the percolators with
a humidity of 80% (w.b.) they are conveyed to
a press where the humidity is reduced to 50%
(w.b.), and then are burned in the boiler
(Dutra, personal communication),
9,12
or alter-
natively, burned after a drying operation to
reduce the humidity to 30%,
3,7
or even to
25%.
5
At these lower levels of humidity, the
burning process in the furnace starts immedi-
ately the material enters the boiler. The re-
duction of humidity to lower levels can carry
a risk of spontaneous combustion.
3,5,7
.
Table 1. Chemical composition of coee ground (dry base)
Authors Protein (%) Gross ®bre (%) Ether extraction (%) Ashes (%)
Adams and Dougan
3
10.8±12.97 37.8±47.6 23.8±29.2 0.27±1.08
Tango
4
12.2 41.0 17.9 4.5
Gopalakrishna Kao and Natarajan
11
14.2 ± 13.9 1.14
P¯uger
12
14.0 37.0 23.0 0.5
Marins
8
15.0 ± 18.0 1.0
Table 2. Heat value of diverse biomass residues
Heat value
Material (kcal/kg dry) (kJ/kg dry) Author
Jute stems 4619.10 19 307.84 Kumar et al.
1
Rice husks 3805.30 15 906.15 Kumar et al.
1
Coconut ®bre 4707.83 19 678.73 Kumar et al.
1
Wheat straw 4185.02 17 493.38 Kumar et al.
1
Cotton twigs 3750.00 15 675.00 Kumar et al.
1
Corn ears 3804.35 15 902.18 Kumar et al.
1
Sugar cane bagasse 4470.59 18 687.07 Anon.
13
Sugar cane bagasse 4600.00 19 228.00 Anon.
13
Wood 5450.00 22 781.00 Anon.
13
Coee grounds 5960.00 24 912.80 Anon.
13
M. A. SILVA et al.458

Approximately 18 kg of coee grounds at
50% of humidity (w.b.) produce the same
quantity of vapour as a gallon of oil.
Therefore, for the typical process of the pro-
duction of soluble coee, 75% of the energy
required by the plant can be supplied.
12
Burning the coee grounds without previous
drying has the following disadvantages: it is
necessary to have a larger combustion
chamber, a surplus of fuel of about 5%
12
and
enough air to maintain the levels of tempera-
ture required. The larger combustion chamber
requires a fan with a greater capacity, and it is
therefore more expensive to burn very humid
coee grounds.
5
Dierent types of dryers can be used to dry
the grounds. Some north-American ®rms use
rotating dryers with internal tubes, another
has introduced a ¯uidized bed to dry from 70
to 25% (w.b.) that is, without previously using
a press.
5,7
Sivetz and Desrosier
7
discuss the
use of pneumatic and ¯uidized dryers, and rec-
ommend the latter. The Brazilian cases will be
discussed below.
In the 1980s coee grounds began to be uti-
lized as a fuel in the soluble coee processing
plants in Brazil. This use began with little
knowledge of the properties of the material
and the best way to burn it. There is a lack of
data on the properties of spent coee grounds
in the literature, and these data are essential
to conduct a study on any one of the pro-
cesses: handling, storage, drying or burning.
The principal purpose of this work is to pre-
sent a qualitative and quantitative analysis
(mass and energy balances) of the coee
grounds burning process as it is carried out in
the Brazilian industry, and a discussion on the
better use of this fuel in boilers, including the
previous drying of the material. Data are pre-
sented, which were collected from question-
naires sent to the plants.
14
The other purpose of this work is to present
some physical and chemical characteristics
that were experimentally determined, the
knowledge of which is necessary before any
study of the previously mentioned process.
These characteristics are: talus angle, apparent
density, real solid density,
15
mean diameter,
super®cial area, heat of combustion and spon-
taneous ignition temperature. Some of these
characteristics are a function of the humidity
of the material. In these cases the study was
conducted with dry and wet material.
16
2. ANALYSIS OF THE BURNING PROCESS
A scheme of the industrial production pro-
cess of soluble coee is shown in Fig. 1.
The roasted and grinding coee is intro-
duced into the percolators in a countercurrent
with the steam, from where exits the ``liquid
coee'', which is introduced in the spray
dryer; spent grounds from the percolators are
introduced into a cyclone to separate the
exhaust vapour and then the liquid water is
removed, using a press and a dryer. Usually
the spent grounds are used as fuel in the boi-
lers.
To obtain the data for the analysis two
types of questionnaires were sent to 11 plants
in Brazil, seven of them answered the ®rst
one, and only ®ve the second one. With these
answers and data from the Brazilian Industrial
Association of Soluble Coee (ABICS),
Tables 3±5 were prepared.
All the plants use oil as an auxiliary
fuel. All of them use a press after the
extraction, and the intermediate transport of
the grounds is made with vapour at high
pressure. Except for plant number 3, all of
them dry the grounds before their intro-
duction into the boiler. The transportation
process after the dryer varies: screw, pneu-
matic, etc.
All the plants use ¯uidized beds to dry the
grounds, except one which utilizes a rotating
dryer. The dryer agent is in all cases exhaust
gases from the boilers. Many dierent types of
boilers and solid fuel feeders are used. In var-
ious cases the boilers are designed to burn
liquid/gas fuels and have been adapted to
burn solid fuels, plant number 2 has a horse-
shoe-type boiler, more appropriate for burning
solid or mixed fuels.
2.1. First law eciency of industrial boilers
To make an analysis of the eciency of the
boilers burning coee grounds, the necessary
parameters were collected from the factories,
and are reported in Table 4.
The energy transferred to steam, in the boi-
ler, was obtained as the dierence of enthalpy
between vapour and liquid water multiplied by
the water ¯ow:
Hws mwhsÿhw
The energy introduced for the fuels was
Biomass residues in the Brazilian soluble coee industry 459

computed, based on their low heat value:
EfmgLHVgmoLHVo
where the subscript gmeans grounds and o,
oil. The mass of grounds considered was dry
mass.
The energy ratio, depending on the source
used, was then calculated as:
rmgLHVg
moLHVo
and the nominal eciency of the boiler, as:
ZHws
Ef
Fig. 1. Soluble coee process ¯owsheet.
Table 3. Information about the type of equipment used in the industry, referring to
dryers, boilers and solid fuel feeders of the boiler
Plant Dryer type Boiler/furnace type Solid fuel feeder
1 Fluidized bed Three-pass ®re tube Conveyor screw
2 Fluidized bed Horseshoe Rotating spreader
3 ± Sloping stationary grate Conveyor screw
4 Rotating rub ± Conveyor screw
5 Fluidized bed Stationary grate Hydraulic piston
6 Fluidized bed Furnace Conveyor screw
7 Fluidized bed Fire tube Conveyor screw
M. A. SILVA et al.460

The values of the enthalpies and the water
vapour saturation temperature were taken
from Van Wylen and Sonntag.
17
To make the
calculation of the combustion reactions, for
the composition of the spent coee grounds
and lower heat value, the values reported in
Silva et al.
16
were adopted. The composition
and heat values of the oils were taken from
Bonomi et al.
18
The speci®c heat of the gases
were taken from Hougen et al.
19
The collected and calculated values corre-
sponding to each of the plants are shown in
Table 4. The eciency of the boilers must be
appraised taking into account that biomass
boilers present values between 65 and 85%,
and oil boilers between 70 and 88%, so, in all
cases, the burning process can be improved.
Considering the data, plant number 2 was the
one with the best values and the data from
plant number 6 are probably wrong.
The unusually low values of eciency are
due to the fact that the boilers are not suit-
able for solid fuel but for liquid or gas fuel,
and have been adapted. In this case they
must use a great quantity of excess air, a fact
that diminishes the eciency; another fact
that also diminishes the eciency is that the
liquid water that constitutes the humidity of
the solid fuel must be evaporated into the
boiler.
2.2. Improvements to the burning process Ð
drying the biomass
One of the most important ways of improv-
ing the burning process is to dry the grounds
with the exit gases of the boiler; this practice
is usual in the soluble coee industry, as was
reported above, but the drying process can be
conducted in a more ecient way in order to
reduce the use of fuel oil. Here we carried out
an analysis of the burning process and the
possible reduction in oil consumption in the
boilers.
From the data obtained (Table 4), some cal-
culations were made, which include a mass
and energy balance. The energy balance con-
Table 4. Data on the burning-steam production process
Parameters/Plants 1 2 3 4567
Soluble coee production (t/day) 5 31 17 ± 28.6 6.1 20
Dry coee grounds (t/day) 6.24 70 18 88 36 6 26
Ratio grounds/fuel (t/t) 1.2 2.3 1.1 ± 1.3 1.0 1.3
Grounds humidity after extraction (%) 87 85 80 80 76 80 85
Grounds humidity after press (%) 55 55 50 55 50 52 55
Grounds humidity after drying (%) 22 35 ± 45 34 40 30
Wet grounds (kg/h) 336 3000 1500 1800 3020 420 1082
Dry grounds (kg/h) 262 1950 750 990 1993 252 757
Auxiliary oil rate (kg/h) 108 500 1000 318 160 630 624
Source energy ratio (kJ/kJ) 1.65 2.65 0.51 2.12 8.47 0.27 0.83
Steam rate (kg/h) 2700 18 000 12 500 6500 10 330 3000 8000
Steam pressure (bar) 16 22 18 15 13 18 21
Steam temperature (8C) 200 216 208 ± 190 ± 214
Steam enthalpy (kJ/kg) 2794.7 2804.8 2799.8 2793.9 2787.6 2798.5 2803.5
Water inlet temperature (8C) 80 98 105 ± 100 ± 80
Water inlet enthalpy (kJ/kg) 335.0 410.7 440.3 104.8 419.2 104.8 335.0
Exit gas temperature (8C) 250 175 260 ± 200 ± 250
Energy transferred to vapour (MJ/h) 6641 43 092 29 493 17 479 24 466 8081 19 748
Energy of fuels (MJ/h) 11 314 72 180 59 692 39 754 59 900 31 679 45 026
Eciency (%) 59 60 49 44 41 26 44
Table 5. Data on the burning process in boilers using grounds and oil as fuels
Parameters/Plants 1 2 3 5 7
Excess of air rate 1.7 3.1 2.7 4.7 2.3
Gas enthalpy H
estg
(MJ/h) 1477 4238 5756 3870 6110
Gas enthalpy H
excg
(MJ/h) 1818 14 557 16 549 22 633 13 669
Gas available heat Q
av
(MJ/h) 1757 4962 8691 7625 9522
Boiler evaporation heat Q
h
(MJ/h) 212 2860 2168 2847 933
Evaporation heat Q
hh
(MJ/h) 919 6490 2168 5525 2657
Oil economy (kg/h) 17.9 91.9 54.8 67.7 44.5
Oil economy (%) 16 18 5 42 7
Biomass residues in the Brazilian soluble coee industry 461

sidered that the chemical energy introduced by
the fuels (E
f
) was used in the form:
EfDHsQhHestgHexcgQL
where the ®rst term on the right corresponds to
the enthalpy of the steam, the second to the heat
spent in the evaporation of humidity in the
grounds, the third to the enthalpy of the exit
gases, considering a burning reaction without
excess air, the fourth to the enthalpy of the
excess air, and the ®fth represents the heat lost.
As the available data were not complete, in
order to do the global balance, two assump-
tions were necessary: (i) the burning of oil and
biomass is complete in the boiler, without the
production of CO; and (ii) heat lost by radi-
ation and convection was considered pro-
portional (10%) to the energy content of the
fuels. The rate of excess air and the enthalpy
of the exit gases are reported in Table 5.
The best energetic option to dry the biomass
is to use the exit gas of the boiler (as it is
being made), but taking advantage of the ther-
mal energy that these gases have: that is their
dierence of enthalpy from the exit tempera-
ture to some adequate ®nal value. This ®nal
temperature must be chosen considering cor-
rosion problems which can occur if the water
vapour in the gases is condensed on the walls
of the dryer; in this study the value considered
was 1008C (a more conservative value might
be 1208C). In Table 5 the values obtained for
the available enthalpy of the gases are
reported as Q
av
.
The values corresponding to the energy
necessary to evaporate the water into the boi-
ler, in actual conditions, and that necessary to
evaporate the water in the grounds at the
humidity at which they leave the press, (Q
hh
),
are also reported in Table 5.
Comparing the ``available'' and the ``necess-
ary'' heat, it can be seen in all cases that it is
higher than that actually used. The two last
columns in Table 5 refer to the potential econ-
omy of oil if the drying process were
improved, in terms of the absolute and relative
quantity of oil actually used today.
3. PHYSICAL AND CHEMICAL PROPERTIES OF
SPENT COFFEE GROUNDS
In order to improve the handling, drying
and burning process of the grounds it is
necessary to know a series of properties of the
material, which were determined and are
reported below.
3.1. Sampling
The samples were obtained using an appro-
priate sample set, repeating the process as
many times as was necessary in order to
obtain the desired sample size. The original
material came from the Nestle
Âfactory located
in the city of Araras, Sa
Äo Paulo, Brazil.
3.2. Sizing
The size of the particles was determined
using a sieve shaker, shaking for 15 min with
a strength of vibration of 7.5 on a scale of 0±
10. The measurement was made with three
samples of approximately 200 g each. The
ABNT standard sieves, numbers 8, 12, 16, 20
and 30, were used; the particles in the bottom
plate were passed once again through sieve
numbers 40, 50, 70, 100 and 140. The mean
ABNT diameter, DpvA, was obtained using the
Sauter mean (volumetric-areolar):
DpvA 
X
k
i1
fmDpiDDpi
rcv
X
k
i1
fmDpiDDpi
rcvDpi
where:
fmDpiDDpi
=retained weighed fraction,
Dpi
=mean diameter of the retained particles.
20
If the density and the shape of the particles
do not depend on the size, then:
DpvA 1
X
k
i1
fmDpiDDpi
Dpi
.
The results are reported in Table 6.
In Table 6 one can see that the ®nest par-
ticles disappear when the degree of humidity
increases; this behaviour is expected because
when the particles become wetter there is
greater adhesion between them, and their size
is increased by the absorption of water. For
this reason, the mean diameter of the particles
increases with humidity.
M. A. SILVA et al.462

3.3. Talus angle
The rest angle was measured using an
appropriate set developed at the Agricultural
Engineering Faculty of the State University of
Campinas (UNICAMP), Sa
Äo Paulo, Brazil.
The material is piled up by means of a con-
stant and free ¯ux of particles. The value in
Table 7 is the mean value of four measure-
ments for each case.
3.4. Apparent/real density and porosity
The apparent density was also measured
using an appropriate set developed by the ®rm
ALEM±MAR (from Brazil). This set has a
container of known volume (1000 cm). This
container was ®lled up by a free and constant
¯ux of particles, then weighed full and empty.
The value in Table 7 is the mean of three
measurements for each case.
The real density of solids
15
was obtained by
the pycnometer method (based on the volume
of a displaced ¯uid). The pycnometer has a
volume of 50 cm. The liquid used was toluene.
21
The value reported in Table 7 is the mean value
of ®ve samples of dry material of 5 g each. The
density of humid material, r
ws
, was obtained
theoretically using the following equation:
rws 1
1ÿuw:b:
rds
uw:b:
rlw
where
r
ds
=dry solid density
r
lw
=liquid water density
In this equation it was assumed that the
volume of the particle changes when it absorbs
water in the same quantity as the volume of
water absorbed.
The porosity was computed theoretically
using the equation:
22
E1ÿrap
rws
3.5. Humidity
The humidity was determined by drying the
material in a drying chamber with forced con-
vection at a temperature of 758C for 24 h. The
value reported in Table 7 is the mean value of
three samples. The number of hours necessary
was previously determined. The temperature
of 758C was used to avoid tar volatilization.
Table 6. Sizing of the spent coee grounds at dierent degrees of humidity
Degree of humidity
(wet basis, %)
Sieve designation Sieve opening 15.23 17.73 43.10
ABNT standard (mm) Retained weighed fraction
±8 2.362 0.0319 0.0303 0.0263
8±12 1.651 0.1349 0.0991 0.1317
12±16 1.168 0.1718 0.1719 0.1745
16±20 0.833 0.2802 0.3032 0.3077
20±30 0.589 0.1078 0.1294 0.2000
30±40 0.417 0.1010 0.1157 0.1284
40±50 0.295 0.0487 0.0659 0.0271
50±70 0.208 0.0353 0.0415 0.0043
70±100 0.147 0.0566 0.0430 ±
100±140 0.104 0.0217 ± ±
140± 0.0106 ± ±
Table 7. Characteristics of spent coee grounds with dierent degrees of humidity
Humidity
(w.b.%)
Talus angle
(o)
Apparent density
(g/cm
3
)
Real density
(g/cm
3
)
Mean diameter
(mm)
Super®cial area
(m
2
/g) Porosity
0.00 ± ± 1.20 ± 5.89 ±
15.23 38.7 0.430 1.16 0.583 ± 0.63
17.73 37.8 0.426 1.16 0.677 ± 0.63
43.10 37.7 0.406 1.11 0.880 ± 0.63
Biomass residues in the Brazilian soluble coee industry 463

In this study the humidity used is in wet basis
obtained as:
uw:b:,%humid weight ÿdry weight
humid weight 100
3.6. Super®cial area
The surface area was determined using an
appropriate set for the measurement of the
surface area of solids, CG 2000, based on the
principle of adsorption of nitrogen on the sur-
face of the solid at its boiling temperature
(ÿ1958C).
This method is based on the work of
Brunauer, Emmet and Teller, who developed
an equation that permits the computation of
the surfacial area through thermodynamics re-
lations of adsorption. This equation is known
worldwide as the BET equation.
15
The technique consists of passing a mixture
of 10% nitrogen in helium ¯owing over a
sample that is cooled at the temperature of
liquid nitrogen at a pressure of 2 atm and rela-
tive pressures below 0.3. Helium is used as a
diluent because it is not adsorbed under these
conditions.
The nitrogen adsorbed at each level of par-
tial pressure changes the composition of the
euent gases. These changes are detected
measuring the electrical conductivity of the
gases through a potentiometric scanner and an
appropriate integrator.
The sample was dried in a drying chamber
and activated at 1008C for 120 min with a
continuous ¯ux of nitrogen of 40 cm/min. The
temperature of 1008C was used because at
higher temperatures tar begins to volatilize.
The talus angle, the apparent density, the
real density and the porosity are reported in
Table 7. The variation of these parameters is
not as signi®cant as the mean diameter, more-
over this small variation is coherent. The
apparent density and the real density decrease
slightly with the humidity and this may be due
to the fact that particles increase in size when
they absorb water, which have less density
than the dry solid. It is known that the poros-
ity depends only on the shape, and not on the
size of the particles, so it remains constant.
The experimental determinations of talus
angle and apparent density are dicult in the
case of the spent coee grounds with 43%
(w.b.) because they are very adhesive and ¯ow
with diculty. For the same reason it is im-
possible to use gravity in order to transport
the material in the factory.
3.7. Terminal velocity
The terminal velocity of the particles, V
t
,
was measured using a specially constructed ap-
paratus based on the principle of ¯uidization.
A sample of the particles is placed on the ap-
paratus, the fan is started and controlled so
that the particles are suspended (¯uidized) by
the ¯ow of air. The air speed is measured with
a Pitot tube. The drag coecient is calculated
from the equation:
CD4
3
Dprsÿrag
raV2
where
V=velocity
r
s
=solid density
r
a
=air density
g=gravity
The experimental drag coecient was com-
pared with values obtained from correlations
for spherical particles reported below.
From Kunii and Levenspiel:
23
CD100:5
Rep
for 0:4<Rep<500
CD0:43 for 500<Rep<200000
From Morsi and Alexander:
24
CD46:5
Rep
ÿ116:67
Re2
p
0:6167 for 10<Rep<100
CD98:33
Rep
ÿ2778
Re2
p
0:3644 for 100<Rep<1000
where the Reynolds number is calculated as:
RepDpraV
ma
where
m
a
=air viscosity
The results are reported in Table 8, and, as
can be expected, the terminal velocity
diminishes with the mean diameter of the par-
ticle (experimental values reported at the sec-
ond column). The fourth column shows the
experimental values for the drag coecient,
which can be compared with the values for
spherical particles with the same diameter and
M. A. SILVA et al.464

Reynolds number, reported in columns 5 and
6, from two dierent correlations.
23,24
The
behaviour of the experimental drag compared
with the values for spherical particles is just
the opposite, for the range tested, that is it
diminishes with the Reynolds number; this is
probably due to a common fact that occurred
in irregular, biological materials submitted to
a grinding process: the density and shape of
the particles present a slight variation with
size.
3.8. Heat of combustion
The heating value of combustion was
obtained through tests in a calorimeter by the
isothermal method adapted to the conditions
of vegetal fuel. The tests were based on the
standard P-MB-454-IBP-ABNT 1968 which
allows corrections to be made to the initial
and ®nal temperature of the bath produced by
little changes in the room temperature. This
standard is equivalent to ASTM-D-240-64.
Corrections must be done because of the nitric
acid formed during the chemical reaction and
also because of the nitrogen that was in the
sample and in the calorimeter at the time it
was set up.
The higher heating value was obtained as
the mean value of three tests of drying
samples. The results, compared with those of
other authors, are presented in Table 7, where
one may observe variations which may be due
to the use of dierent types of coee (dierent
blends) in the process, or to dierences in the
extraction process. It is interesting to point
out that the heat value of spent coee grounds
is similar to that of charcoal and is higher
than that of ®rewood.
11
The carbon±hydrogen±nitrogen analysis was
made by the Chemical and Chemical
Engineering Division of the Technological
Research Institute (IPT), Sa
Äo Paulo, Brazil
(see Table 9).
Knowing the quantity of hydrogen in the
sample, it is possible to compute the low heat-
ing value subtracting the change of enthalpy
between vapour and liquid water. The value
for dry and humid material is shown in
Table 10. Should it be necessary, the low heat-
ing value for wet samples can be obtained as:
LHV 25240 ÿ276
uw:b:,%kJ=kg wet material
where u(w.b.,%) is the humidity of the
sample, in percentage, wet basis.
3.9. Spontaneous ignition temperature
The tests made were based on the work of
Hardman et al.,
25
who studied the spon-
taneous combustion of an activated coal bed.
Several tests, using a crucible with a thermo-
electric joint, were made in a tubular electric
oven with a controlled heating rate. A con-
stant air ¯ux was blown into the oven in order
to provide oxygen for the combustion. After
the same preliminary tests, the ®nal essays
were made with a rate of heating of 58C/min
and an initial temperature of 2008C (in the
oven).
It was observed that the start of the volatil-
ization occurs at a temperature of approxi-
mately 1958C, with the liberation of black
smoke and a coee smell. At 2608C, the tem-
perature increment rate, recorded by the ther-
moelectric joint, increased as if self-ignition
were taking place, although this rate returns
quickly to the normal value of the oven.
Finally, at 4608C, spontaneous ignition occurs,
which was made evident by the increase in the
temperature rate. The temperatures reported
before varied by 258C between the tests. It
was impossible to reduce these dierences.
Table 8. Experimental terminal velocity and drag coecient of coee grounds
D
p
V
t
C
D
C
D
C
D
(mm) (m/s) Re
p
Coee grounds Spheres
23
Spheres
24
2.01 5.1 616 1.1 0.43 0.52
1.41 4.1 353 1.1 0.53 0.62
1.00 3.9 238 0.9 0.65 0.73
0.71 3.2 138 0.9 0.85 0.93
0.50 2.9 89 0.8 1.06 1.12
Table 9. Carbon±hydrogen±nitrogen
analysis of the spent coee grounds
Carbon (C) 59.5%
Hydrogen (H) 7.3%
Nitrogen (N) 2.5%
Oxygen (O) 30.7%
Biomass residues in the Brazilian soluble coee industry 465

The data on the spontaneous ignition tem-
perature must be considered for future prac-
tice uses such as storage, drying or
combustion. These values depend on the quan-
tity of oxygen and the size and humidity of
the material, so they must only be considered
as indicative values. In the case of the storage
of wet material, fermentation can occur with
formation of methane. A complete investi-
gation of this subject would be very useful, to
avoid accidents.
4. CONCLUSIONS
Data on the industrial use as fuel of spent
coee grounds in the Brazilian soluble coee
industry, are reported.
Compared with other biomass, coee
grounds can be considered an excellent fuel,
because they have a heat of combustion higher
than others usually used. With some techno-
logical eort to improve the conditions of util-
ization of this fuel in the boilers, the soluble
coee industry could become self-sucient in
thermal energy. Some investment would be
required, in cases where the type of boiler or
dryer is not adequate.
To improve the process of storage, drying
and burning of spent coee grounds, some
indispensable data on the characteristics of the
residue are necessary, and these data were
obtained and reported in detail.
AcknowledgementsÐThe authors would like to thank
Prof. Elisabeth Jorda
Äo (FEQ/UNICAMP) who made the
determination of the super®cial area possible.
REFERENCES
1. Kumar, K., Bal, S. and Ojha, T. P., Fuel character-
istics of agricultural residues, Agricultural
Mechanisation in Asia, Africa and Latin America,
1984, 15, 61±64.
2. Jenkins, B. M. and Summer, H. R., Harvesting and
handling agricultural residues for energy, Transactions
of ASAE, 1985, 29, 824±835.
3. Adams, M. R. and Dougan, J., Waste Products,
Coee, VII Ð Technology, ed. R. J. Clarke and R.
Macrae, Elsevier Applied Science Publishers, London,
1985, pp. 291.
4. Tango, J. S., Utilizac°a
Äo industrial do cafe
Âe dos seus
subprodutos 28, 49±73, 1971.
5. Sivetz, M., Coee Processing Technology, Vol. 2, AVI
Publishing Co., Wesport, 1963.
6. Clarke, R. J. and Macrae, R., Coee, Vol. 1 Ð
Chemistry, Elsevier Applied Science Publishers,
London, 1985.
7. Sivetz M. and Desrosier, N. W., Coee Technology,
AVI Publishing Co., Wesport, 1977.
8. Marins, J. M., Anais do Semina
Ârio sobre Biomassa
como Energia na Industria, 1, Rio de Janeiro, 1980,
pp. 12.01±12.04.
9. Landais, C., Charbon et marc de cafe
Â,Revue de
l'Energie, 1986, 384, 341±389.
10. L. A. H. Nogueira and L. F. V. Flores, Potencialidade
energe
Âtica de Borra de Cafe
Âpara as Industrias de
Cafe
ÂSolune. Pesquisa e Desenvolvimenton Tecnolo
Âgico
13, 15±25, 1987.
11. Gopalakrishna Kao, N. and Natarajan, C. P.,
Utilisation of coee by products, Indian Coee, 1974,
38, 3±5.
12. P¯uger, R. A., Solid Wastes: Origin, Collection,
Processing and Disposal, ed. C. L. Mantell, Wiley
Interscience Publishers, Toronto, 1975, pp. 365±376.
13. Anon, X., Let residue disposal pay for itself, Power,
1971, 115, 60±61.
14. Machado Silva, M. J., Silva, M. A. and Nebra, S. A.,
Utilizac°a
Äo da Borra de Cafe
Ânas Fa
Âbricas de Cafe
Â
Solu vel no Brasil. In Proceedings of the VIth
Congresso Brasileiro de Energia e I Semina
Ârio
Latinoamericano de Energia, 18±22 October, Rio de
Janeiro, Brazil, 1993, pp. 771±777.
15. Gregg, S. J. and Sing, K. S. W., Adsorption Surface
Area and Porosity. Academic Press, London, 1982,
303 pp.
16. Silva, M. A., Sanchez, C. G. and Nebra, S. A.,
Physical and Chemical Properties of the Residue of
the Soluble Coee Industry. In Proceedings of the Ist
Asian-Paci®c International Symposium on Combustion
and Energy Utilisation, 15±18 October, Beijing, China,
1990, pp. 125±129.
17. Van Wylen, G. J. and Sonntag, R. E., Fundamentos de
la Termodina
Ãmica Cla
Âssica, ed. Edgar Blucher Ltd.,
Sa
Äo Paulo, Brazil, 1978.
18. Bonomi, A., Lima, A. F., Pinto, J. H., Sant'anna, R.
J., Hochgreb, S. and Park, S. W., Conservac°a
Äode
Energia na Indu stria de Papel e Celulose. Instituto de
Pesquisas Tecnolo
Âgicas, Sa
Äo Paulo, Brazil, 1985.
Table 10. Heat of combustion of spent coee grounds (kg/kg dry) at dierent degrees of
humidity (w.b.)
Authors 0% 25% 50% 75%
Tango
4
5000
Gopalakrisna Kao and Natarajan
11
5000
Sivetz and Desrosier
7
5000 4830 4440 3330
P¯uger
12
5560 5330 4940 3720
Adams and Dougan
3
6930
Marins
8
5345
Anonymous
13
5960
This paper HHV 6415
LHV 6031 5841 5462 4324
M. A. SILVA et al.466

19. Hougen, O. A., Watson, K. M. and Ragatz, R. A.,
PrincõÂpios de los Procesos QuõÂmicos, ed. S. A. Reverte
Â,
Spain, 1964.
20. Foust, A. S., PrincõÂpios das Operac°o
Äes Unita
Ârias, ed.
Guanabara Dois, Rio de Janeiro, 1982.
21. Mohsenin, N. N., Physical Properties of Plant and
Animal Materials: Structure, Physical Characteristics
and Mechanical Properties. Gordon and Breach
Science Publishers, New York, 1970.
22. Shubert, H., Food particle technology: Part I: proper-
ties of particles and particulate food systems, Journal
of Food Engineering, 1987, 6, 1±32.
23. Kunii, D. and Levenspiel, O., Fluidization Engineering,
ed. Robert E. Krieger, Huntington, 1977, 534 pp.
24. Morsi, S. A. and Alexander, A. J., An investigation of
particle trajectories in two-phase ¯ow systems, Journal
of Fluid Mechanics, 1972, 55, 193±208.
25. Hardman, J. S., Street, P. J. and Twamley, C. S.,
Studies or esontaneous combustion imbeds of acti-
vated carbon, Fuel, 1980, 59, 151±156.
Biomass residues in the Brazilian soluble coee industry 467