Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~Z373~
Case 2969
DESCRIPTION
CONTROLLED COFFEE ROASTING
Technical Field
The present invention relates to a co~ee
05 roasting method and particularly, to a cofEee roas-ting
method permitting control of final product properties.
Background ~rt
Conventional non-fluidized bed coffee roasting
equipment uses a roasting atmosphere at a temperature
oten exceeding 540C. In addition, conventional
non-fluidized bed roasters typically use low weight
ratios of roasting atmosphere to coffee beans (herein-
after referred to as the air to bean ratio). The
combination of the high roasting temperature and low
air to bean ratio results in a temperature gradient
from bean to bean as well as within the coffee bean
itself. Such a temperature gradient indicates that
different roasting reactions occur at different
times among the beans as well as internally in any
given coffee bean.
Fluidized bed roasting of coffee beans is well
established in the art. For e~ample, U.S. Pat.
No. 4,169,164 to Hubbard et al. describes a two stage
fluidized bed roasting process wherein the temperature
in the irst sta~e is between about 226C and 243C
and the temperature in the second stage is between
- 2 - ~ ~3~3~
268C and 285C. A similar, improved process is
disclose~ in U.S. Pat. No. 4,322,447 to Hubbard
wherein ~he temperature in both stages is between
287C and 299C although the velocity of the roasting
05 atmosphere is varied between the two stages. The
apparatus contemplated for use in both disclosures
is a fluidized bed apparatus wherein heated gas is
directed downwardly through jets onto a vibrating
gas-impervious plate which gas is then deflected
upward, thereby fluidizing the roasting coffee
beans. The apparatus is more fully described in
U.S. Pa~. No. 3,229,377 to Hoyt. Numerous modii-
cations oE the apparat~ls are disclosed in U.S. Pat.
Nos. 3,262,217, 4,109,394, 4,169,322, 4,201,~99 and
4,306,359. Both the Hubbard and Hubbard et al.
processes operate for at least a portion of the
roast at temperatures well in excess of 240C and
hence, fail to gain the advantages of the present
invention.
Another apparatus for the fluidized bed roasting
of coffee at temperatures not in excess of 276C is
disclosed in U.S. Pat. No. 3,964,175 to Sivetz.
Again, by roasting at temperatures as high as 276C,
the Sivetz apparatus does not offer the unique
advantages of the present invention. The Sivetz
disclosure also contains an extensive survey of the
prior art attempts at fluidized bed roasting. The
processes described therein are unlike the present
invention.
It is an object of the present invention -to
provide a coffee roasting method which permits
better control of the final product properties.
It is another object of the invention to provide
a roast:`~ng method which provides greater control
over the roasted whole bean density.
1;2~732~
-- 3
It is another object of the invention to provide
a roasting method to produce a less dense coffee
having flavor strength and soluble solids yield
equal or better than that of a denser conventionally
05 roasted coffee.
It is a further object of the presPnt invention
to provide a roasting method with greater process
control, owing to the lower roasting temperatures
used therein.
Brief Description of the Drawing
Figure 1 is a representational plok of the
results obtained by individually roasting a number
of coffee beans in a thermomechanical analyzer. The
figure indicates the expansion of the coffee bean as
a function of the bean temperature during roasting.
Detailed Disclosure of the Invention
It has now been found that the objects of the
invention are met by a coffee roasting method using
a single roasting atmosphere temperature at or less
than about 240C throughout the entire roasting
cycle wherein coffee beans are suspended in a bubbl-
ing bed in a gas heated to a temperature selected
from the range between 200C and about 240C and
maintained that way for from 2 minutes to 10 minutes.
The roasted coffee beans are then discharged and
cooled.
A key feature of the present invention is
controlling the temperature to which the roasting
coffee i<, exposed to a temperature selected from the
range between 200C and about 240C. It is known
that the so-called pyrolysis reactions, those
reactions responsible for the characteristic coffee
flavor, aroma and color, begin to occur only at
~237~;22
-- 4 --
temperatures in excess of about 185C as described
in Coffee Processin~ Technology, Sivetz and Foote,
Westport, Conn., AVI Publishing, Vol.l p.230, (1963).
Exposing coffee beans to tempera-tures in excess of
05 about 240C leads to surface charring or burning of
the beans, increasing roasting losses and impairing
-the flavor of the coffee so roasted. By limiting
the temperature of the roasting atmosphere to one
selected Erom the range between 200C and about
240C in the present method, the coffee beans are
heated -to a -temperature sufficiently high so as -to
induce p~rolysis but low enough -to avoid damaging
the bean" -thereby providing a superior roasted
coffee.
It has also been found that flavor and color
develop~llent reactions occur uniformly among the
beans and within each coffee bean if the roasting
atmosphere is maintained at a temperature at or
below about 240C. Structural stresses within the
coffee beans are reduced, permitting greater control
over roasted whole bean density. The desired
temperature uniformity as well as rapid heat transfer
rate are achieved by suspending the coffee beans in
a bubbling bed with a constant temperature roasting
atmosphere maintained below about 240DC.
Maintaining the roasting atmosphere within the
prescribed range is also essential for the reason
that it is within said range that the so-called
glass transition temperature, Tg, is reached. The
glass transition temperature, as defined in polymer
chemistry is that temperature at which a polymer
loses its crystalline structure, softens and becomes
amorpho~ls, like glass. A coffee bean is known to be
comprised of a large portion of crystalline carbo-
hydrates, such as mannan and cellulose. When Tg is
3~3~
-- 5 --
reached, the carbohydrates contained in said beanssoften and lose -the crystalline s-tructure. Internal
pressure produced by gases generated during roasting
causes the softened mannan and cellulose within the
05 bean to expand, "opening up" the bean to roasting
and providing the density decrease that is an object
of the present invention.
The phenomena of a coffee bean passing through
its glass transition temperature is observed from
Figure 1. Figure 1 is a representative plot of the
txpical results obtained from roasting cofEee beans
in a thermomechanical analyæer. Such a -thermo-
mechanical analyzer measures the expansion or con-
traction of a material about a given axis in response
to a controlled heat input. Thermomechanical analyzers
are often used in the polymer industry to determine
Tg, which is indicated by the changed slope of the
expansion curve measuring the expansion about a
given axis. As Tg is reached, the polymer (in tnis
case, mannan and cellulose) softens and the internal
pressure causes the markedly different, nearly
vertical, rate of expansion observable on the expan-
sion curve. In Figure 1, the representative shape
of which was obtained by roasting coffee beans
individually in a Perkin-Elmer Thermomechanical
Analysis System ~, Tg is seen to occur between about
210C and 225C as indicated by the changed slope of
the curve in the region marked "A".
The roasting atmosphere temperature is not
limited to within the range of 200C to 240C in
conventional roasting and in fact, said temperature
is deliberately chosen to be substantially higher
than 240C. 'IConventional roasting" refers to
roasting in typically commercial equipment, such as
a Thermalo roaster manufactured by Jabez Burns &
~3732~
-- 6 --
Sons, Inc., wherein relatively low velocity roasting
atmosphere contacts an essentially static bed of the
coffee beans at inlet temperatures of between 370C
and 560C. The roasting atmosphere temperature is
05 often modulated to within the range of 340C to
530C toward the end of the roast. Flame time, that
is, the amount of time the hot roasting atmosphere
is actually circulated through the roaster, is
usually 10 -to 12 minutes, with the total roasting
cycle taking about 15 to about 17 minutes. A faster
conventional roasting technique is disclosed in U.S.
Pat. No. 4,349,573 -to Stefanucci e-t al. wherein the
roastins atmosphere temperature is increased to
about 640C with a modulation temperature between
365C and 520C. The flame time is correspondingly
reduced to between 5 and 7.5 minutes. It is empha-
sized t.l~at conventional roasting uses an essentially
static bed, heat transfer to which is relatively
slow and non-uniform.
Roasting within the prescribed temperature
range of 200C to 240C affords greater control over
the roast color of the coffee in comparison to
conventional roasting. In conventional roasting,
the temperature of the roasting coffee beans is
constantly increasing, without ever reaching or
indeed, closely approaching the roasting atmosphere
temperature hereinbefore described. Thus, the rate
of the roasting reaction is constantly accelerating,
making the point at which the roast is terminated
critica:l, for if the roast proceeds even 10 seconds
longer than targeted for instance, the roasted
coffee will be significantly darker than desired.
In the roasting method of the present invention, the
temperal:ure of the coffee beans rapidly approaches
the temperature of the roasting atmosphere so that
_ 7 _ ~23~
the rate of the roasting reaction is fairly steady.
A relaively small variation in the roast time will
not cause the roasted coffee -to be significantly
darker, particularly when a roasting atmosphere
05 temperature toward the lower end of the specified
range is used.
Another important feature of the present inven-
tion is roasting in a bubbling bed, which bubbling
bed promotes rapid and uniform heat transfer from
the roasting atmosphere to the roasting coffee
beans. A "bubbling bed" is intermediate be-tween a
static, non-fluidized bed wherein essentially none
of the beans are suspended in the roasting atmosphere
and a fluidized bed wherein substantially all of the
beans are suspended in the roasting atmosphere.
Bubblinq bed as used herein is one in which the
greater mass of coffee beans is suspended in the
roasting atmosphere at any given time, with a smaller
mass of beans comprising a static bed. There is
constant circulation within the bubbling bed, with
any given coffee bean being suspended for between
about 50% and 70% of the time. The rapid heat
transfer promoted in a bubbling bed causes the
temperature of the beans to closely approach that of
the roasting atmosphere, minimizing the thermal
gradient between said beans and roasting atmosphere.
It has been discovered that a minimal thermal gradient
is desirable in furthering uniformity among the
roasting coffee beans as well as minimizing structural
stresses in the bean, permitting optimal bean expan-
sion (and hence, an optimal roasted whole bean
density decrease).
The bubbling bed is formed by suspending -the
greater mass of coffee beans in a large volume of
upwardly flowing roasting atmosphere. One difference
- 8 - ~ ~373~Z
between conventional static bed roasting and bubbling
bed roasting is shown by the weight ratio of roasting
atmosphe~e to coffee beans, the so-called air to
bean ratio. In a typical conventional commercial
05 coffee roaster, the air to bean ratio is about 1.0
kg air (roasting atmosphere)/1.0 kg roasting cofee
beans. The air to bean ratio of the present method
though, is preferably from 40.0 to 150.0 kg air/1.0
kg coffee beans. With the high air to bean ratio,
the individual coffee beans are surrounded by an
"envelope" of roastiny atmosphere, contributing to
the uniEormity of the roasted coffee. The air to
bean rat~io for a fluidized bed is between 10.0 and
30.0 kg air/l.0 kg coEfee beans. The inlet veloci-ty
of the roasting atmosphere needed to maintain a
bubbling bed is also intermediate between the velocity
in a conventional roaster and the roasting atmosphere
velocity in a fluidized bed roaster. In a bubbling
bed roaster, the velocity is on the order of 670
m/min to 1250 m/min compared to between 40 m/min and
46 m/min for a conventional roaster and on the order
of 3660 m/min for fluidized bed roasting. As is
apparent, the maintenance of the bubbling bed as
compared to a fluidized bed differs primarily in the
inlet velocity of the roasting atmosphere.
The heated gas or roasting atmosphere which is
used to suspend and maintain the bubbling bed may be
indirectly heated air or preferably, air combined
with the combustion gases (principally carbon dioxide
and water) of -the heat source, typically a burner.
Combining the air wi-th -the combustion gases is
preferred because of the greater efficiency of such
an arrangement. Additional energy efficiency is
achieved by recirculating a majority of the roasting
atmosphere throughout the roasting cycle. Recircu-
12~73~
g
lation is particularly convenient in the present
~ inventiorl because the lower temperature used herein
elimina-tes the surface burning of the coffee beans
and hence, smoke formation associated with con-
05 ventional roasting.
Manipulation of the roasting atmosphere temper-
ature within the prescribed range, as well as of the
roasting time between about 2 minutes and 10 minutes
provides sufficient control so as -to produce roasted
coffees with varying properties. Table 1 illustrates
the different roasted whole bean densities attainable
(at nearly constant roast color) by varying the
roasting temperature and time. The roasted whole
bean density is seen to decrease steadily with
increasing temperatures and correspondingly shorter
roasting times. Table 2 illustrates the different
roasted whole bean densities attainable by varying
the roasting time at constant roasting temperature.
Roasted whole bean density is seen to decline with
increasing roasting time. In addition, it is seen
that the density of the coffee produced by the
present method may be made either greater or lower
than that obtainable by conventional coffee roasting
(densities between about .30 gm/cc and .50 gm/cc for
the method of the present invention compared to
about .32 gm/cc for conventional roasting).
Such improved density control is a signiicant
advantage of the present invention, greatly expand-
ing the variety of roasted coffee products that may
be produced. For example, the ability to produce a
higher density product is useful in soluble coffee
processing because a greater weight of said coffee
can be loaded in an existing percolator column,
increas ng productivity. The ability to produce a
lower density coffee provides a product which product
-10-
~;~3~32~
C rl
~o
~ o~ a
~ ~ ,~ $ O
a ~ ~
~4
C
U
o, ~ _, ~00 0 ~ O ~ ~ C~ 00 0 ~ 00 t` ~o
~ o In ~ 001~ l O ~ c~
o O r~
o ~ Ul ~~r O ~00 o ~ ~ ~ ~ U`~
O U~ ~u~ U~ ~u~ ~ u~ ~D ~O 00 00 00 00 00
C
U ~
C ~ o ~r7 ~ o o
E C C Cr~ C C C C -~ C C C C ~1 C
o E e E e E E E o E E E E o E
o
C
ooooo ooooo ooooo
o ~ ~ ~ ~ o ~ ~J ~ ~ o C`l ~ ~ ~
o ~ ~I N c~ L'~ ~O
E i
,~ g 2 o o
C C C C C C
,~
o D o D C ~ C
~; I ~ U ~
~,~ `,1
s~
C S~
~,~ E
C ~ O ~ :~
~237~
.~
~ C~ ~ _~ oo ,~ ~D1` ~ ~ ~ ~ '~ '~
oo~~ o~ U~ ~ ~ ~C~ oo ~ o
Kal
o
C o o c~ u~ ~ O O L~ D O ~ ~) ~
K A A ~ A
I E~
C " e
o
.~ ~
~ O ~ 0~
~J
.,1 ~rl
C X D
1:1 O N
1:~ ~
~;2373;~
Ll
C
u, C O ~ ~ ~ ~ X O ~ o~ ~1 ~ O In u~
~ 7 ~ o oo r~
o
o o oo ~ ~ g g oo ~ o g ~ Uo~ ~
O~ A ~ ~ ~- _1
,_
0
C E
,~,, E~ Q~ aJ
~_ ~4~ ~ ~J
~ C C C
kl
~ l
~0
C
o ~ o o
C .Q
.
~Q
~n ~ ~
.,~ ~ ~
C I'd O D
;I~ S.l ~ O
a!~
~;~3~2~
- 13 -
permits using a lesser weight of coffee to give the
same cup strength and as much or more aromatic
coffee flavor. The lower density of the product
enables the consumer to use the same volume of
05 coffee in the preparation of said coffee, which
volume of coffee weighs less than an e~ual volume of
a more typically dense conventionally roasted coffee.
The consumer realizes a cost savings in using less
coffee without altering the customary recipe level.
Although a lesser weight of the lower density
coffee may be used, a brew prepared from said coffee
is as strong as a brew prepared from a greater
amount clf a conventionally roasted coffee. Table 3
shows that both the soluble solids and the flavor
strength ~perceived by an expert panel) are greater
for a coffee produced by the present invention as
compared to a conventionally roasted coffee. The
roasted coffee prepared by the method of the present
invention was roasted in a bubbling bed at a temper-
ature of 232C for about 2.5 minutes. The convention-
ally roasted coffees were roasted in a commercial
unit wherein heated air is blown through a perforated
rotating cylinder containing the coffee beans to
conductively heat the cylinder and convectively heat
the beans. The inlet air temperature was between
about 370C and 560C for lO minutes for the slow
conventional roast and at a temperature of approxi-
mately 640C for 5 minutes for the fast conventional
roast.
~73Z~
- 14 -
TABlE 3
Roasted CoEfee Roast Color Recipe % Soluble Perceived Strength
cups/lb Solids Scale of 5-10
Milds
05 bubbling bed 54 90 23.9 5.9
conventional (fast) 48 90 l9.5 5.3
conventional (910w) 52 75 21.3 5.0
Brazils
bubbling bed 63 90 25.1 6.1
conventional ~East) 62 90 24.4 5.7
conventional (slow) 63 75 21.5 5.0
Robustas
bubbling bed 83 90 23.0 5.2
conventional (fast) 82 90 23.0 5.3
conventional (slow) 82 75 . 18.6 5.0
The brew prepared at 75 cups/lb. was made with
63.4 gm. coffee/1780 ml. water and the brew prepared
at 90 cups/lb. was made with 51.2 gm. coffee/1780
ml. water. Perceived strength of the brew is based
on a scale of 5 to 10 with the slow conventionally
roasted coffee being assigned a base value of 5
in each case.
-
~,3~3~
- 15 -
The roasting method of the present invention
may be either batch-wise or continuous. In a batch-
wise scheme, the green coffee beans are loaded into
the roaster. The roasting atmosphere is then heated
05 to a temperature within the specified range between
200C and about 240C and circulated upwardly through
the bed of coffee beans at a velocity between 670 m/
min and 1250 m/min in order to create the bubbling
bed. The air to bean ratio is between 40.0 and
10 150.0 kg air/1.0 kg coffee beans. After the beans
have been roasted for a sufEicient period o time,
the circulation of the roasting atmosphere is halted.
The roasted coffee beans may then be rapidly cooled
to a temperature below 65C by either chilled air or
water ~uenching.
A continuous roasting method is preferred
because of the relative ease of operation and greater
uniformity obtainable with a continuous method. The
preferred scheme is one in which the green coffee
beans are continuously fed to a rotating perforated
cylinder contained in the roasting apparatus. The
cylinder is compartmentalized by a helical screw
contained therein. As the cylinder rotates, the
coffee beans are moved for~ard from one compartment
to the next by said rotating helical screw. The
beans charged to one compartment remain separated
from the others, preventing any mixing of the coffee
beans a-t different degrees of roast. Such separation
insures that essentially all of the coffee beans
have the same residence time within the roaster,
which residence time is controlled by the rate of
rotation of the perforated cylinder. The roasting
atmosph~ere, preferably comprised of the combustion
gases oE the heat source, is forced upwardly along
the whole length of the perforated cylinder with
~L23~22
- 16 -
sufficient force and at a sufficient volume to
create b~Lbbling beds of the beans contained in each
compartment. Thus, the air to bean ratio and inlet
roasting atmosphere velocity are most preferably as
05 described for the batch bubbling bed roaster. The
rotating action of the perforated cylinder provides
further advantageous agitation within the bubbling
beds, in addition to that provided by the upwardly
flowing roasting atmosphere. The roasted coffee
beans are discharged rom the apparatus and prefer-
ably cooled to a temperature below 65C by air or
water quenching.
The continuous bubbling bed roas-ting method of
this invention is preferable to a fluidized roasting
process wherein the beans are fluidized by impinging
a gas on a gas impervious plate such as is disclosed
in U.S. Pat. No. 4,169,164 to Hubbard et al. or U.S.
Pat. No. 4,322,447 to Hubbard. The fluidized bed
roasting method makes no provision for preventing
back-mixing within the fluidized coffee bed. Conse-
quently, coffee beans charged to the roaster at the
same time do not necessarily have the same r2tention
time therein, with some beans exiting sooner and
others exiting later than those charged at the same
time. The result is, of course, lessened uniformity
of the roasted coffee. The continuous bubbling bed
method though, uses the helical screw which in
effect compartmentalizes the roaster, insuring the
uniformity of the retention time of coffee beans
charged at a given time. Such uniormity of reten-
tion time leads -to desirable uniformity of the
roasted coffee so produced.
The following examples illustrate certain
embodiments of -the present invention. The examples
are not intended to limit the invention beyond what
is claimed below.
~:3~3;~
- 17 -
Example 1
1. 454 gm of green Colombian Mild coffee having
about 10% by weight moisture were placed in a "V"
shaped perforated basket and placed in a batch Ross
05 Dryer manufactured by Midland-Ross Co. of New srunswick,
New Jersey.
2. The coffee beans were then roasted at 232C with
a roasting atmosphere velocity of 1220 m/min which
flow rate was sufficient to suspend the beans
in a bubbling bed. The air-to-bean ratio was about
52 kg air/kg bean.
3. The roast was terminated after 2 min. 45 sec.
by shutting off the burner and removing the basket
from the Ross Dryer.
4. The roasted beans were rapidly cooled by a
forced flow of air at ambient temperaturP. The
results for the roast are shown below.
roast color: 50 color units
roasted whole bean density: .332 gm/cm3
The roast color was determined by grinding a
sample of the coffee, screening out the fine fraction
and compressing the same in a Carver Press to form a
tablet and determining the light reflec-ted from said
tablet in relation to an arbitrary standard as
measured by a Photovolt detector. The method of
measuring roast color is fully described in Goffee
Processing Technolo~y, Sivetz and Foote, Westport,
Conn, A~I Publishing, Vol. 2 pp. 132-137.
~23~
- 18 -
The roasted whole bean density was determined
by the so-called free flow method wherein roasted
whole coffee beans are poured through a funnel into
a vessel of known volume. The vessel and coffee are
05 weighed, the vessel tare weight subtracted and the
density calculated therefrom.
The roasted coffee was then ground and a brew
prepared therefrom. The brew was characterized by
an expert panel as being woody, aromatic and acid,
with a strong cllp strength.
Example 2
1. Green Colombian Mild coffee beans were fed to a
Ross Helical Suspended Particle Dryer (HSP) also
manufactured by Midland-Ross, Co. The Helical
Suspended Particle Dryer is one wherein the beans
are fed to a perforated rotating cylinder having a
helical screw therein. The rotating screw advances
the green coffee therethrough. The roasting atmosphere,
a combination of the burner combustion gases and hot
air, is blown upwardly along the bottom length of
the cylinder at a rate sufficient to suspend the
beans in a bubbling bed. The cylinder used for this
example had a 41% open area with 3/32" diameter
holes on 9/64" centers.
2. The coffee beans were fed to the ~ISP a~ a rate
of 16 kg/hr. The perforated cylinder was rotating
at a rat.e of 3 RPM giving a roast time of 2 min. 30
sec. The inlet roasting atmosphere was at 238C and
was blown through the bubbling bed at a velocity o
about 1()00 m~min.
- 19 - ~2~37~22
3. The beans were rapidly cooled by a forced flow
of ambient air after being discharged from the HSP.
The results for the roast are shown below.
roast color: 46 color units
05 roasted whole bean density: .321 gm/cm3
The roasted coffee was then ground and a brew prepared
therefrom. As in Example 1, the brew was characterized
by an expert panel as being woody, aromatic and
acid, with a strong cup strength.
