FLEXGAS PROJECT

Deliverable DL2.4
Procedure for the determination of the fuel behaviour under more realistic conditions of
gasification

Upon feeding into a hot fluidized bed, a fuel particle undergoes a series of phenomena as shown in the
graph of Fig. 1. Each step is described hereinafter along with suggestions for determining the particle
behaviour under nearly realistic conditions as well as some experimental data obtained for fuels of
interest for the FLEXGAS project.

Char
Drying Devolatilisation Ash
conversion

Comminution

Fig. 1: stages occurring during FB conversion of a fuel particle

1. Particle drying
It is the first step of the whole conversion process for the fuel particle. A slight overlap with the
devolatilisation step occurs at the late stage of drying.
Drying takes place very quickly, the moisture being normally under 20% by mass for practical fuels. A
rough estimation of the drying time is given by Eq. 1. It is assumed that the sub-process is mainly
controlled by the thermal exchange between the fuel particle and the dense bed needed for the
vaporization of the water at 100°C.

X mV p ρ p λv
td = Eq. 1
H (Tb − 100 )Ap

where
td = drying time
Ap = particle external surface
Vp = particle volume
Xm = moisture content
Tb = bed temperature
H = coefficient of heat transfer
λv = heat of water vaporization
ρp = particle density
Table 1 reports the values of td for different fuel particles of interest in the FLEXGAS project.
Reasonable values of properties/parameters are reported in the table. In summary, the particle drying
lasts a few seconds, confirming the little relevance during the whole particle conversion.

Tab. 1: drying time of four different fuels (estimated values)

Fuel Density Size Tbed H Moisture Drying time

kg/m3 mm °C W/(m2K) s

German brown 860 3 800 200 0.150 1.04

coal
Polish coal 1190 4 800 200 0.070 0.90
Wood pellets 1250 6x20 800 200 0.084 2.21
Olive bagasse 1350 4 800 200 0.089 1.29

The direct measurement of the drying time under realistic conditions is not easily viable. However,
some information could be obtained by adapting the experimental procedure proposed by Daugaard &
Brown [1] for determining the enthalpy of pyrolysis of fuels in fluidised bed.

2. Particle devolatilisation
The devolatilisation occurs after particle drying, as shown in Fig. 1. Since the devolatilisation is an
endothermic step, the volatile matter is released from the particle thanks to the heat transferred from the
surrounding environment (e.g. fluidised bed). Therefore, the enthalpy of devolatilisation/pyrolysis [1]
is a characteristic parameter that depends on the fuel nature as well as on the particle heating rate, and
in turn on its geometrical properties. For fluidised bed application, it can be evaluated in a standard
equipment as described by Daugaard & Brown. Values in the range 0.8-1.5 MJ/kg are reported for
different biomass fuels (i.e. oak, pine, corn stover).
The direct observation of the particle behaviour in fluidised bed allows to have information on the
devolatilisation time and the number of bubble (Nb) issuing from the fuel particle. In fact, the release of
fuel volatiles, also including tar species, gives rise to a train of so called “endogenous” bubbles [2]
during devolatilisation. These bubbles rise the bed and erupt at its surface.
A devolatilisation time of around 30 s was measured from the direct observation of flames from wood
and coal-wood pellets in lab-scale combustors. For instance, Fig. 2 reports some frames obtained by
video-recording the bed surface (120 mm ID) during devolatilisation of a coal-wood pellet (frame 1)
locked in the bottom of the bed by means of a cage, allowing to measuring the devolatilisation time.
Bubble eruptions appear in frames 3 and 5, whereas a flame is clearly shown in frame 6.
 Fig. 2: frames obtained by video-recording the bed surface during devolatilisation of a single pellet

Experimental values of Nb, in the range 2-12, at temperature of 1123 K are reported by Fiorentino et al.
[2] for three fuels, namely wood chips, bituminous coal and TDF. However, Nb largely depends on fuel
properties, particle size and bed temperature. The indirect measurement of Nb can be obtained by
analysing and working out the pressure time-profile obtained by immersed probes in a bed at minimum
fluidisation, since it is affected by the passage of bubbles.
It is likely that also under gasification conditions, the devolatilisation behaviour is similar to
combustion; in fact the volatiles form a shield that surrounds the fuel particle and prevents any
chemical reaction with external species. (e.g. O2, CO2, H2O), independent of the oxidising/reducing
conditions of the environment where the particle is.

3. Char conversion
Relevant information on char conversion can be obtained by thermo-gravimetric analysis.
An example of an experimental procedure adopted at Istituto Ricerche sulla Combustione – CNR [3] is
proposed hereinafter. Char samples can be prepared in fixed bed tubular furnaces with heating rate up
to 50°C/min and temperature up to 1540°C or in bubbling fluidised bed reactors at 850°C. Fractions in
the size range <300µm are used to prepare char samples in heated strip reactors under severe heating
conditions: temperature up to 2000°C, pressure up to 12 bar, heating rates up to 105°C/min. After
pyrolysis char samples are ground and sieved to the size range <150µm for reactivity study by TG
analysis.
Thermogravimetric experiments are carried out in isothermal or non isothermal mode using a
thermobalance Netzsch 409C which features maximum temperature of 1500°C and is coupled through
a skimmer device with a Mass Spectrum (0-300amu) and with IR CO/CO2 analyzers.
In non isothermal experiments samples are dehumidified in the TG for 5 min at 100°C in an upward
flow of 150ml/min. The temperature is then raised from 100°C to 900°C at a constant heating rate, HR,
comprised between 5 and 100°C/min under gasification conditions. The sample is finally held at 900°C
for 30min. In isothermal experiments of char gasification the char is heated up in the thermobalance to
a given temperature TR in a flow of nitrogen with HR of 50°C/min. The gas is then switched from
nitrogen to the desired reactive atmosphere and isothermal reaction takes place till carbon burn-out. In
experiments of char gasification with carbon dioxide the reaction temperature TR is comprised between
500 and 100°C and CO2-N2 mixtures are used with carbon dioxide partial pressure in the range 0.02-
1bar.
It must be emphasized that in all the experiments the amount of sample loaded is kept very small
(generally around 20 mg and sometimes even below 5mg) and particles are evenly spread on the
sample holder. These precautions are necessary to avoid problems of particle transfer.

Analysis of thermogravimetric data

The mass loss recorded during non-isothermal experiments is worked out to obtain the derivative of the
mass loss as a function of temperature (DTG curves). Comparison of the peaks of the DTG curves is
used to gain a quick reactivity index of different materials.
The differential TG analyses for Polish coal and German brown coal are shown in Fig. 3. Three peaks
correspond to water evaporation, volatiles release and char conversion. German Brown coal presents
larger mass losses upon water evaporation and devolatilisation as shown in Fig. 3. Furthermore, the
peak of char conversion is significantly shifted toward lower temperature, around 800 °C, when
German brown coal is considered. It clearly appears that the German brown coal, denoted by the red
curve, is much more reactive in CO2 gasification than Polish coal (blue curve).

2,5
DTG (mass % /min)

1,5

0,5

0
0 200 400 600 800 1000

Temperature ( °C)

Fig. 3: Differential TG analyses of Polish coal ( --- ) and German brown coal ( --- ) in CO2 atmosphere,
heating rate 5 oC/min

4. Particle comminution

In the fluidized bed fuel particles are also subject to comminution phenomena that influence the
particle size distribution and produce elutriable fines. In particular, the loosely connected or even
incoherent structures of chars left behind by devolatilisation, leading to a more pronounced impact of
particle attrition phenomena, dramatically change the fate of char. Characterisation of the impact of
these phenomena during fluidized bed combustion/gasification is of great importance for the
 determination of the reaction rate and
residence time of fuel particles as well as for
the evaluation of conversion efficiency loss
due to small unconverted particles of char,
called carbon fines, escaping the reactor.
Comminution of fuel particles taking place in a
fluidized bed combustor can be seen as the
result of four phenomena occurring in series or
parallel with each other and with
combustion/gasification (Fig. 4).
Primary fragmentation is that occurring,
immediately after the injection of the fuel
particles in the bed, as a consequence of
thermal stresses caused by rapid heat up of the
particles and by internal overpressures due to
volatiles emission. The degree of
fragmentation is strongly affected by the type
of fuel (volatile content, swelling index and
mechanical resistance) and by fuel particle size
and shape.
Fig. 4. Series-parallel comminution phenomena During their residence time in the fluidized
(Chirone et al., 1991). bed, char particles are also subject to
secondary fragmentation and attrition by
abrasion. Both processes are determined by mechanical stresses due to collisions with other particles or
with the internals of the reactor. These phenomena are classified on the basis of the typical size of the
generated fragments: secondary fragmentation generates coarser and substantially non-elutriable
fragments; attrition by abrasion generates finer and quickly elutriable fragments. Secondary
fragmentation depends on the resistance of the particles, whose structure is progressively weakened by
reaction, to impacts against walls and internals in the bed or in the jetting region. Attrition by abrasion
should rather be related to the resistance of the particles to surface wear. Attrition by abrasion can also
strongly depend on the reaction conditions in the bed: internal reaction and uneven progress of reaction
promote the increase of particle voidage and, in turn, the decay of the mechanical strength of the
carbon material. The enhancement of the fines generation rate due to parallel particle burning has been
called “combustion/gasification-assisted” attrition. However, coarse char particle residence times in the
gasifiers are expected much longer than those relative to fluidized combustors and a direct consequence
is that attrition phenomena during gasification assume a much more relevant impact on carbon
conversion efficiency.
The fourth phenomenon, percolative fragmentation, is associated to the loss of connectivity of the char
porous structure which occurs when particle voidage increases, at high conversion degrees, beyond a
threshold. It results in the generation of either coarse or fine fragments, depending on the porous
texture of the solid.
The elutriation is the mechanism by which part of the bed material, whose size is to a first
approximation such that the particle terminal velocity is lower than the superficial gas velocity in the
bed, is carried away with the flue gases. Another important mechanism in the freeboard of a fluidized
bed is particle entrainment: bed particles, whatever their size, are ejected into the freeboard by bubbles
bursting on the bed surface. Since most bed material is non elutriable, particles rapidly fall back to the
bed. Non elutriable particles during their fall back to the bed carry with them some small particles
impacted in their trajectory. Also clustering phenomena and other particle–particle or particle–wall
interactions have been observed in the freeboard.
Primary and secondary fragmentations usually do not participate in the generation of elutriable fines.
The loss of carbon from the bed is due only to elutriation of fines entering with the coal feed, fines
formed by attrition, fines resulting as residues from combustion/gasification of original particles (in
general a limited quantity) and fines formed, when it occurs, by percolative fragmentation. Primary and
secondary fragmentations, however, indirectly affect combustion efficiency by influencing the particle
size distribution and in turn the extent of carbon surface exposed to abrasion and the number of
particles present in the bed.
Procedures - Typical procedures for the characterization of the comminution behaviour and
determination of comminution properties are reported hereinafter [4]. A stainless steel atmospheric
bubbling FB combustor 40 mm ID and 1 m high was used for fragmentation and attrition experiments.
The temperature of the bed, measured by means of a chromel-alumel thermocouple placed 40 mm
above the distributor, was kept constant by a PID controller driving an electric furnace. The freeboard
is kept unlagged in order to minimize fines post-conversion in this section. Gases were fed to the
column via high-precision digital mass flowmeters.
Two different reactor configurations were used for the tests. In the first configuration (Fig. 5A), used
for particle fragmentation experiments, the top section of the fluidization column was left open to the
atmosphere. A bed of sand (0.3–0.4 mm, 180 g) was fluidized with nitrogen at 0.3 m/s. A stainless steel
circular basket can be inserted from the top in order to retrieve fragmented and unfragmented particles
from the bed. The tolerance between the column walls and the basket is limited to reduce as much as
possible the amount of carbon left in the bed when pulling out the basket. The basket mesh is of 0.8
mm, so that the bed sand can pass through the net openings. During the run a stainless steel circular
basket was inserted from the top to retrieve the particles from the bed and lied on the distributor
without interfering with the bed fluid dynamics. Following the procedure proposed by Chirone et
al.(1991), experiments were carried out by injecting single fuel particles into the bed kept at 800°C
from the top of the column. After about three minutes, required to completely devolatilize the fuel, the
resulting char was retrieved by means of the basket in order to investigate the number and size of the
produced fragments. The experiment was repeated with more than 30 particles in order to collect a
statistically significant number of fragments.
 AIR/CO2

A B
1) gas preheating section; 2) electrical furnaces; 3) ceramic
insulator; 4) gas distributor; 5) thermocouple; 6) fluidization
column; 7) steel basket; 8) manometer; 9) digital mass
flowmeters; 10) air dehumidifier (silica gel).
1) gas preheating section; 2) electrical furnaces; 3) ceramic
insulator; 4) gas distributor; 5) thermocouple; 6) fluidization
column; 7) head with three-way valve; 8) sintered brass
filters; 9) hopper; 10) SO2 scrubber; 11) stack; 12) cellulose
filter; 13) membrane pump; 14) gas analyzers; 15) personal
computer; 16) manometer; 17) digital mass flowmeters; 18)
air dehumidifier (silica gel).

Fig. 5: experimental apparatus used for comminution tests

In the second configuration (Fig. 5B), used for fines elutriation rate (attrition) experiments, a two-exit
brass head was fitted to the top flange of the column. By operating a valve it was possible to convey
flue gases alternately to two removable sintered brass filters. Batches (2 g) of pre-devolatilized fuel
char were fed to the bed (0.3–0.4 mm sand, 180 g). The bed was kept at 800°C and fluidized at 0.8 m/s
either with nitrogen (inert conditions) or with a N2-CO2 mixture with 60% CO2 concentration
(gasification conditions). Elutriated fines were collected by means of the two-exit head by letting the
flue gas flow alternately through sequences of filters (one was in use while the previous one was
replaced) for definite periods of time. In order to prevent hydration of the collected material, each filter
was readily put in a drier after being used, where it was cooled down before it was weighed. The
difference between the weights of the filters before and after operation, divided by the time interval
during which the filter was in operation, gave the average fines elutriation rate relative to that interval.
Fines collected in the filters during the fuel char tests were further analyzed to determine their fixed
carbon content. The assumption underlying this procedure was that the residence time of elutriable
fines in the reactor could be neglected and that elutriation rate could be assumed equal to the rate of
fines generation by attrition at any time.
Experimental results - Table 1 reports an overview of the results of primary fragmentation experiments
for three sample fuels: commercial spruce wood pellets; home-made pellets composed either of 70%
pine wood – 30% German brown coal (GBC) or 70% pine wood – 30% Polish bituminous coal (PC).
The starting materials for the home-made pellets were all in the particle size range 0.5-2.0 mm. All the
pellets had cylindrical shape with a diameter of 6 mm and an average length of 20 mm.
 Table 2: Results of pelletised fuel primary fragmentation tests.
Fuel D0, mm Sf n1 d1, mm
Spruce-wood pellets 7.27 0.20 1.75 5.96
70% wood - 30% German coal pellets 7.35 0.21 2.47 6.37
70% wood - 30% Polish coal pellets 7.30 0.30 3.30 6.24

Results are expressed using the following quantities: D0 = Sauter mean diameter of the initial pellets; Sf
= primary fragmentation probability, the number of fuel pellets that undergo fragmentation divided by
the total number of pellets fed; n1 = primary fragmentation multiplication factor, representing the
number of generated fragments per unit fuel pellet fed to the reactor; d1 = Sauter mean diameter of the
char particles after devolatilisation. Results show that for all fuels limited primary fragmentation
occurs, with a fragmentation probability between 20 and 30% and particle multiplication factor of the
order of 2-3. The slight decrease of the Sauter mean diameter upon devolatilisation is caused by the
combined effect of primary fragmentation and particle shrinkage. These results are confirmed by the
cumulative particle undersize distributions (on mass basis) of the char pellets after devolatilisation as
compared to the initial fuel pellets size distributions, which showed that the population of smaller
fragments (< 5 mm) does not significantly increase.
Figure 6 shows the rates of carbon fines elutriation measured during batch attrition experiments with
the three pelletised fuel chars. The carbon elutriation rate (EC) was normalized with respect to the
initial amount of fixed carbon fed to the reactor (WC0). Results of experiments carried out under inert
and gasification conditions (60% inlet CO2 concentration) are compared. The curves under inert
conditions are similar one to the other and follow the typical decreasing trend towards an asymptotic
value. However, the extent of fines attrition is different among the three chars. Spruce-wood char
appears to be harder than the other two, suggesting that the commercial pelletisation procedure yields
more compact pellets than the home-made one. This is reflected by the total amount of elutriated
carbon upon the runs: the percentage of the initial carbon fed with the fuel char that was cumulatively
elutriated during the 120 min runs in inert condition was 7, 21 and 12% for the spruce-wood, German
coal – wood and Polish coal – wood chars, respectively.
 A completely different behaviour was noted under
gasification conditions. All the curves showed a
significant peak in the carbon elutriation rate at some
point during particle conversion. It is worth to note
that under gasification condition the tests lasted until
complete carbon consumption, as witnessed by the
vanishing of the CO concentration at the FB outlet. As
a consequence, the total duration of the runs reflected
the intrinsic reactivity of the fuels. As expected, the
spruce-wood char was the most reactive, while the
Polish coal – wood char was the least. The presence of
the peak in the carbon elutriation rate indicates that
carbon consumption in the pellets progressively
weakens the char structure by pore enlargement, much
like the well known combustion-assisted enhancement
mechanism typically observed during FB coal
combustion. This gasification-assisted enhancement is
dependent on the parent fuel reactivity, since lower
rates of carbon consumption shift towards longer times
the peak position. The decreasing part of the curves at
the right of the peaks are obviously caused by the
progressive decrease of the carbon loading in the bed
that overtakes the attrition enhancement at late stages
of carbon conversion.
The percentage of the initial carbon fed with the fuel
char that was cumulatively elutriated upon the runs in
gasification condition was 14, 29 and 38% for the
spruce-wood, German brown coal – wood and Polish
coal – wood chars, respectively. This figure shows that
the carbon loss by elutriation is certainly one of the
Fig. 6: Results of fuel char attrition tests. critical factors during the gasification process,
especially for the char produced from the less reactive
Polish coal. On the other hand, the higher reactivity
and mechanical strength of the commercial spruce-wood pellets appear to be beneficial for obtaining a
larger carbon conversion during gasification.

References
1. Daugaard D.E., Brown R.C., Enthalpy for pyrolysis for several types of biomass, Energy &
fuels 17, p. 934-939 (2003)
2. Fiorentino M., Marzocchella A., Salatino P., Segregation of fuel particles and volatile matter
during devolatilization in a fluidized bed reactor. II: Experimental, Chem. Eng. Sci., 52, p.
1909-1922 (1997)
3. Senneca O., Kinetics of pyrolysis, combustion and gasification of three biomass fuels, Fuel
Processing Technology 88, p. 87-97 (2007)
4. Chirone R., Massimilla L., Salatino P., Comminution of carbons in fluidized bed combustion,
Progr. in Energy and Combust. Sci. 17, p. 297-326 (1991).