Fuel 79 (2000) 793802

www.elsevier.com/locate/fuel

Conversion of coal particles in pyrolysis and gasification: comparison of

conversions in a pilot-scale gasifier and bench-scale test equipment
Y. Zhuo, R. Messenbock, A.-G. Collot, A. Megaritis 1, N. Paterson*, D.R. Dugwell, R. Kandiyoti
Department of Chemical Engineering and Chemical Technology, Imperial College (University of London), London SW7 2BY, UK
Received 7 December 1998; received in revised form 2 August 1999; accepted 4 August 1999

Abstract
Comparison with bench-scale experiments has been used to evaluate strategies by which fuel conversion could be improved in a pilot-scale
air blown fluidised bed gasifier. Bench-scale data clearly showed that lower overall conversions observed in the pilot scale gasifier do not
correspond to levels of reactivity expected from the properties of the original feedstock. Increasing particle sizes were found to affect
conversions adversely. Conversions also depended on the reactivity of the gaseous medium steam CO2 and increased with reactive gas
pressure. Relative combustion reactivities of residual chars diminished rapidly with time-at-temperature: at 1000C, char reactivities
declined sharply between 0 and 10 s holding time. The adverse effect of increasing pressure and of secondary char deposition on char
reactivity were experimentally significant but small. In low temperature gasifiers, the combined effect of longer times required for consuming
larger char particles and the rapid deactivation of chars by (10 s) exposure to gasification temperatures tends to lead to appreciable levels of
residual char removal. Our findings indicate that coal/char reactivities decrease appreciably during the first 10 s at peak temperature
underlining some the difficulties of establishing reactor simulations, which include fundamental kinetic models of coal gasification. The
kinetics of char deactivation as a function of time-at-temperature and particle size would appear to play an important role in determining
eventual gasification conversions as well as the gasification kinetics itself. 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Gasification; Char reactivity; Particle size; Reactor design

1. Introduction
Air blown gasification constitutes one of the basic process
stages of the new generation of cleaner coal technologies
being developed in the UK and in the USA. The Air Blown
Gasification Cycle [1] pilot plant (ABGC) constructed by
British Coal (Stoke Orchard, UK), employs a spouted bed
gasifier operated at approximately 970C and pressures up
to 25 bar. The design of the spouted bed air/steam gasifier
has been optimised to operate at a particular combination of
gas throughput, particle residence time and fluidising velocity that requires a 3 mm topsize feed. Estimates of conversion by steam-gasification in the gasifier are low and range
between 5 and 20% (probably closer to the lower value); the
added steam was primarily thought to act as a heat sink to

This paper was presented at the IEA Fluidised Bed Conversion Workshop on Single Particle Conversion in Fluidised Beds (Paris 12 November
1998).
* Corresponding author.
E-mail address: n.paterson@ic.ac.uk (N. Paterson).
1
Present address: Department of Mechanical Engineering, University of
Birmingham.

control the temperature of the combustion section and to

avoid ash melting.
During pilot plant tests, between 2025% of the feed coal
was removed from the gasifier as coarse and fine char
and combusted in a Circulating Fluidised Bed Combustor
(CFBC) to produce steam. The reactivity of the char recovered from the gasifier bed was low, which was thought, at
first glance, to have limited the conversion that was attained
in the gasifier. Increasing fuel conversion in the gasifier
improves overall cycle efficiency, since a greater proportion
of energy in the coal is converted through the gas turbine
and waste heat boiler, which is more efficient than the steam
turbine alone.
There appears to be good reasons for linking the formation of relatively large amounts of unreactive char in pilot
scale air blown gasifiers, to the use of relatively large feed
coal particle sizes. In this study, comparison with benchscale experiments has been used to investigate whether
higher levels of fuel conversion are possible at comparable
temperatures when smaller particle sizes are used. In addition to pressure and particle size, the dependence of conversion and char reactivity on the heating rate and the holding
time at peak temperature have been investigated. Possible

0016-2361/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved.
PII: S0016-236 1(99)00194-5

794

Y. Zhuo et al. / Fuel 79 (2000) 793802

Table 1
The reactor operating conditions
Parameter

Wire mesh

Fluidised bed

Fixed bed

Coal
Gaseous environment
Heating rate (K s 1)
Hold temperature (C)
Hold time (s)
Sample size (mm)
Sample weight (mg)
Pressure (bar)

Daw Mill (washed)

He/Steam/CO2
1000
1000
10
1000
60
106152 (and 600850 micron in the FBR)
6
50
50
130

mechanisms underlying the effect of fuel particle size on

levels of tar release and char reactivity have been briefly
reviewed. The implications of the results for understanding
reactor parameters, which determine fuel behaviour in pilot
and larger scale fluidised bed gasifiers have been briefly
discussed.
In conducting this study care was taken to ensure that the
char preparation conditions in the bench scale tests replicated as closely as practical, the conditions of the larger
scale gasifier. This was because earlier studies have demonstrated that the reactivity of the generated char is very
dependent on its conditions of formation. For this reason it
is not valid to extract meaningful data from the literature to
predict the reactivity of chars derived from air blown gasifiers, because of their unrepresentative conditions of formation. Although the wealth of data in the literature does give
an insight into the subject of char reactivity and helps to give
an understanding of the process science [2].

2. Experimental
2.1. Laboratory scale reactors
A wire mesh reactor and a fluidised bed reactor were used
for the majority of the experimental work of this study. They
were chosen as they could be operated with heating rates,
residence times, temperatures and pressures that were
considered to be relevant to pilot and larger scale air
blown gasifiers. A fixed bed reactor was used to generate
chars at low heating rate, as examples of unreactive material. Experiments using other types of reactor to generate
char, such as drop tubes or TGA were not considered, as
they were not able to operate under relevant conditions.
The development of the wire mesh and fluidised bed
reactors has been described in detail in several publications
[37]. Conditions used in the reactors during the present set
of experiments have been summarised in Table 1.
2.1.1. The wire mesh reactor
The coal sample is placed as a mono-layer between two
layers of folded mesh, which is resistance heated between
water cooled electrodes. Gas enters through the base of the
reactor and is directed through a flow smoothing section into

the mesh. Two thermocouples are attached to the mesh for

lateral temperature measurements and temperature control.
The reactor is contained within a pressure shell capable of
operating at up to 150 bar. Temperatures up to 1500C can
be reached at heating rates variable between 1
10,000 K s 1.
The reactor has recently been modified for carrying out
steam gasification experiments [5], where steam condensation on internal surfaces has been avoided by restricting and
preheating the flow-path of the steam used as sweep gas.
Timetemperature plots from steam injection experiments
in the new reactor configuration have shown levels of
temperature stability comparable to those observed during
experiments with dry gases.
2.1.2. The fluidised bed reactor
The reactor consists of a INCOLOY 800 HT tube
(500 mm tall and 34 mm internal diameter), designed to
withstand pressures up to 40 bar at 1000C. It is lined
with a loosely fitting quartz tube to limit corrosion and
potential catalytic effects of the metal casing. The current
configuration uses a fluidised sand bed (weight 30 g),
heated to reaction conditions before the batch addition of
the coal sample charge. The static bed height is approximately 35 mm. Weight losses are determined from differences between initial and final combined weights of bed
solids and residual carbonaceous material. Chars are recovered from the bed by density separation for subsequent
characterisation work. A holding time of 60 s is used for
obtaining repeatable data [6].
2.1.3. Fixed-bed (hot-rod reactor)
The reactor consists of an INCOLOY 800 HT (or NIMONIC 105: Henry Wiggin Alloys) tube, 250 mm tall and of
6 mm internal diameter, designed to withstand pressures up
to 40 bar at 1000C. A fixed-bed of coal is heated from
ambient to the required temperature, whilst a gas stream
of variable composition sweeps through the reactor. The
maximum heating rate that can be achieved, whilst maintaining an acceptable radial temperature distribution
(10 K s 1) is much lower than that of the wire mesh and
fluidised bed reactors [7].
2.2. Char reactivity
Many types of reactor have been used to prepare chars for
reactivity measurement, such reactors include TGA, drop
tubes, wire-mesh reactors, fluidised beds and fixed beds.
Several techniques have been used to assess the reactivity
of the chars produced and these include measurement of
surface area and combustion reactivity. Surface area
measurement requires surface areas to be in excess of
100 m 2 g 1, which generally requires a sample weight in
excess of 20 mg. This is greater than the weight of char
produced in WMR tests and this has prevented the use of
the technique in this study, although it has been widely used

Y. Zhuo et al. / Fuel 79 (2000) 793802

795

Table 2
The analysis of Daw Mill coal
Proximate (% wt, daf)

Ultimate (wt%, daf)

Maceral

Volatile matter,

Ash,

O (by diff)

Vitrinite

Liptinite

Inertinite

39.9

4.7

80.1

4.7

1.3

1.1

12.8

66

13

21

by other workers [8]. The isothermal, relative char

combustion reactivity has been used as the measure of
char reactivity in this study, as this requires only a low
sample weight. Values were determined by a standard
TGA test (e.g. cf. Ref. [5]), carried out at 500C in air.
Under these conditions the reaction is thought to take
place under chemical control. Results are generally
expressed as Rmax, the maximum rate of weight loss,
although selected data is also expressed as t1/2, the
time required for 50% burnout. The repeatability of
the Rmax determination is ^9% of the value quoted.

2.3. The fuel

The laboratory standard Daw Mill coal sample (Table
2) has been used in most of the bench-scale experiments
described below. Some bench scale experiments have
also been carried out using the sample of Daw Mill
coal from the pilot plantwhich had a higher ash
content. Neither sample melts during slow heating at
atmospheric pressure; however, some melting is
observed during heating under pressure and at high
heating rates.

3. Results and discussion

3.1. Gasification in steam
Fig. 1 presents conversions (total volatile yields)
observed during steam gasification of Daw Mill coal in
the three bench-scale reactors. All experiments were
conducted with 60 s holding at peak temperature
(1000C). As in the case of gasification in CO2 [9,10]
conversions monotonically increased with pressure.
However, using 20 bar steam nearly complete conversion
of the sample was obtained in both the wire-mesh and the
fluidised bed reactors. These two reactors have been specifically designed to segregate individual coal particles from
one another and operate at high heating rates: agreement
between them was within experimental error. In line with
results from previous hydropyrolysis and CO2-gasification
experiments [911], lower conversions were obtained in the
fixed bed (hot-rod) reactor. In steam, as well as in less
reactive CO2 (see below), there appears to be an upper
limit to conversion for reaction times of the order of 60 s
at 1000C. Char reactivity data presented below strongly
suggest these lower conversions to be related to lower heating rates used in the fixed bed reactor (10 K s 1, allowing

Total Volatile Yield in Steam, %, daf

100

80

60

40

Wire Mesh, steam

Fluidised Bed, steam
Hot Rod, steam

20

0
0

10

15

20
Pressure, bar

Hold Time 60s

Fig. 1. Weight losses in steam in the three reactors.

25

30

35

796

Y. Zhuo et al. / Fuel 79 (2000) 793802

60
Wire Mesh, He
Fluidised Bed, He
Hot Rod, He

Total Volatile Yield in He, %, daf

55

50

45

40

35

30
0

10

15

20

25

30

35

Pressure, bar
Hold Time 60s
Fig. 2. Weight losses in He in the three reactors.

long exposure times at temperatures near 1000C), as well

the previously discussed [9,11] poor gassolid contact due
to stacking of softened coal particles within the fixed-bed.
Conversions reported in Fig. 1 correspond to the sum of
weight loss by pyrolytic devolatilisation and weight loss due
to reactions between carbonaceous material and steam. It is
possible to determine the extent of sample weight loss due
to gasification alone, by deducting the weight loss observed

by pyrolysis, during experiments performed under otherwise identical conditions. Fig. 2 presents weight loss data
from experiments in helium, carried out in all three reactors,
using the same set of experimental conditions as in Fig. 1. In
all three reactors, the weight loss decreased with increasing
pressure; the effect is well known and associated with
suppression of volatile release by the physical effect of pressure
[12]. It appears that the re-deposition and re-polymerisation of

60

Extent of Gasification, %, daf

50

40

30
Steam Gasif WMR
Steam Gasif FBR

20

Steam Gasif HRR

CO2 Gasif WMR

10

CO2 Gasif FBR

CO2 Gasif HRR

0
0

10

15

20

25

Pressure, bar

Hold Time 60s

Fig. 3. Comparison of extents of gasification in steam and CO2.

30

35

Y. Zhuo et al. / Fuel 79 (2000) 793802

797

Table 3
Combustion reactivity of short hold time chars from the WMR and ABGC chars (repeatability of the Rmax TGA measurement is ^9% of the value given)
Source of char

Hold time (s)

WMR, He

WMR, CO2

10

ABGC, bed
ABGC, cyclone
ABGC, filter

Variable

10

Pressure (bar)

Temperature (C)

Reactivity (Rmax)

1
10
20
30
1
10
20
30
13

1000

2.53
2.35
2.35
2.29
1.09 (0.89 a)
0.89 (0.62 a)
0.91 (0.69 a)
0.93 (0.59 a)
0.31 a
0.42 a
0.58 a

1000

970

1.90
1.18
1.03
1.01

Char derived from the high-ash Daw Mill coal sample used at CTDD.

tar precursors on char surfaces lead to loss of tar yield and

the formation of less reactive secondary char [10,13].
In pyrolysis, agreement between the wire-mesh and the
fluidised bed reactors was within experimental error over
the pressure range (130 bar). As in the case of steam gasification, the data reflect similar high heating rates and
substantial particle segregation achieved in both reactor
configurations. By contrast, lower tar and total volatiles
were observed in the hot-rod reactor. As previously
discussed [11,14], this appears to be due to the configuration
of the reactor, which allows re-deposition and re-polymerisation of tars whilst passing through the fixed bed of heated
solids. Furthermore, at low pressures, less tar is released
from coal particles due to the lower heating rate
(10 K s 1) [14]. The combined effect is visible in Fig. 2 at
pressures between 1 and 10 bar.
However, above 10 bar, higher heating rates tend to
reduce tar and total volatile yieldsfor coals which soften
at least partially [3,4]: Daw Mill coal softens when heated
rapidly and under elevated pressures. Fig. 2 shows that,
above 10 bar, volatile release in the two high heating
rate reactors drops more rapidly with increasing
pressure.
It may also be noted that greater differences were
observed between conversions in the fixed-bed and the
other two reactors during steam gasification (Fig. 1)
compared to pyrolysis (Fig. 2). Once the appropriate
temperature range is reached, pyrolytic volatile release
from coal particles appears to be less inhibited by effects
due to particle stacking, compared to gasificationwhich is
more directly dependent on gassolid contact.
Fig. 3 presents extents of steam gasification calculated
by subtracting weight loss during pyrolysis (Fig. 2) from
total conversion in steam-gasification (Fig. 1). Experiments
carried out in the fluidised bed reactor with 1/1 mixtures of
steam and CO2 gave results that were nearly indistinguishable from conversions in pure steam. For completeness,
corresponding extents of CO2-gasification (from Refs.
[5,9]) have been included.

In all reactors, the extent of gasification in steam was

markedly greater than that in CO2. However, char agglomeration during CO2-gasification above 20 bar has been difficult to control and this was reflected in the lower conversion
observed in the fluidised bed at 30 barcompared with the
result obtained in the wire-mesh reactor. Measures adopted
to control agglomeration in the fluidised bed reactor have
already been described in Ref. [6]; effects due to agglomeration were not found during operation with steam and
were less pronounced during the pyrolysis experiments. In
the fixed-bed (hot-rod) reactor, extents of steam gasification above 10 bar, were also found to be somewhat above
the upper limit of estimated levels of conversion by steamgasification in the pilot-plant gasifier. Smaller extent of
gasification values were found in CO2.
To summarise, data in Figs. 1 and 3 show that, above
10 bar, conversions and extents of gasification determined
in the wire-mesh and the fluidised bed reactors were quite
significantly higher than total conversions and estimated
steam gasification levels in the pilot-plant gasifier. Results
from experiments in CO2 reflected the same trend although
values were lower. The data clearly show that the lower
overall conversions observed in the pilot scale gasifier do
not correspond to levels of reactivity that may be expected
from the properties of the original feedstock (Daw Mill
coal) itself.
3.2. Char combustion reactivities
Table 3 presents char reactivities determined by the TGAbased method outlined above; results were expressed in
terms of the maximum rate of weight loss, Rmax. Data for
steam gasification chars was limited due to high conversionsand little available char residuefrom wire mesh
and fluidised bed reactor experiments.
The ABGC chars were produced from power station
grade Daw Mill coal, with a higher ash content than the
washed coal used as the laboratory standard sample. Both
Daw Mill coals have been tested; the reactivity of chars

798

Y. Zhuo et al. / Fuel 79 (2000) 793802

Table 4
Combustion reactivities of pilot plant chars from five different coals
Coal

Bed char
Primary cyclone fines
Rmax (% min 1, daf)

Candle filter fines

Bed char, crushed

Daw Mill (14% ash)

El Cerrejon
Reitspruit
Illinois No 6
Drayton

0.31
0.25
0.17
0.51
0.30

0.58
0.35
0.35
0.66
0.43

0.34
0.24
0.25
0.43
0.22

0.42
0.25
0.38
0.52
0.53

from the high ash Daw Mill coal was generally about 2/3 of
that of the low ash coal [15]. Topsize coal (3 mm) was
used in the pilot-scale gasifier and three char streams were
produced: (i) from the base of the gasifier; (ii) primary fines
caught in the primary cyclone immediately downstream of
the gasifier; and (iii) candle (dust) filter fines collected
downstream of the cyclone.
Table 3 shows that reactivities of the ABGC chars varied
between streams. At the fluidising velocities encountered in
the pilot gasifier (typically 0.8 m s 1), the finer particles
(70 mm) are thought to be rapidly removed by pneumatic
transport whilst larger particles apparently spent minutes (or
hours) in the bed prior to removal with ash. Thus the bed
charsmostly larger particleswere the least reactive;
when particles were crushed before reactivity tests, differences were found to be within experimental error. The fines
were more reactive as these probably represent less aged
material, containing some finer coal that was rapidly
elutriated from the reactor. The candle filter fines were the
finest material and had the highest reactivity. This is thought
to be a result of their finer size, and hence larger surface
area.
Several further points emerge from the data in Table 3.

(ii) At 1000C, char reactivities declined sharply between

0 and 10 s holding time, whilst the effect of increasing
pressure was experimentally significant but small.
(iii) The three chars from the ABGC pilot plant were all
considerably less reactive compared to chars prepared in
the wire-mesh reactor from 105152 mm coal particles.
Table 4 shows similar low reactivities for all five coals
used in the pilot gasifier test programme.
Fig. 4 presents changes in the combustion reactivities of
chars from 1000C, atmospheric pressure CO2 gasification
experiments as a function of heating rate (1 to 10,000C s 1)
for three holding times (0, 10 and 60 s). The reactivities of
the 0-s holding chars were observed to change rapidly with
heating rate. The low values encountered for chars from
slower heating rate experiments result from long exposures
to higher temperatures during the slow heatup. However, the
data clearly shows a very sharp drop in reactivity and an
almost total absence of heating-rate dependence above 10 s.
Presumably the chars continue to lose reactivitybut more
slowlywith increasing time at temperature.
Table 5 presents combustion reactivities of 60-s chars
from pyrolysis, steam- and CO2-gasification experiments
at 1000C, showing that:

(i) Pyrolysis chars from the wire-mesh reactor were more

reactive than CO2-gasification chars.

(i) As above, pyrolysis chars were found to be marginally

Comb. Reactivity Rmax [ % / min, daf ]

3.5

0s

10 s

3.0

60 s
2.5

2.0

1.5

1.0

0.5
1

10

100

1000

10000

Heating Rate [ C/s ]

Fig. 4. Effect of heating rate and holding time at 1000C on the combustion reactivity of Daw Mill chars.

Y. Zhuo et al. / Fuel 79 (2000) 793802

799

Table 5
The test programme and results for the combustion reactivity tests (chars produced from Daw Mill coal at hold temperature: 1000C; hold time: 60 s; particle
size range: 106152 mm)
Pressure (bar)

1
10
20
30
1
2
10
20
30
1
10
20
30

Gaseous environment

He

Steam

CO2

Combustion reactivity, Rmax (% min 1, daf)

Wire mesh char

Fluidised bed char

Hot rod char

1.26
1.0
0.77
0.71
1.36

0.30

1.02
0.86
1.02
1.04

1.6
0.8
0.8
0.8

1.1
1.1
1.1
1.1

1.2
1.2
1.0
1.0

0.37
0.24
0.24
0.15
1.0
1.0
0.9
0.9

more reactive than chars produced in gasification experiments.

(ii) Overall, reactivity values were lower compared to
those of shorter residence time (0, 10 s) chars shown in
Table 3.
(iii) A downward trend of reactivities with pressure could
still be observed although differences were smaller than
those of shorter holding time chars shown in Table 3;
some differences were within the band of experimental
scatter.
(iv) Amongst all cases shown in Tables 3 and 5, the
steam-gasification chars were the least reactive and
comparable to reactivities of chars from the pilot-plant
gasifier.
(v) In accordance with Fig. 4 showing the chars to be
largely dead after the first 10 s, the differences between
chars from the three reactors were not very marked. At
longer times, effects due to the slower heatupand
longer exposure to high temperaturestaking place in
the fixed-bed (hot-rod) reactor were barely observable.
3.3. Effect of particle size on pyrolysis and gasification
behaviour
The large particle fraction of the 3 mm top size feed to
the pilot-plant appears to provide most of the unreacted
Table 6
The effect of coal particle size in the fluidised bed reactor (coal, Daw Mill
(washed); temperature, 950C; pressure, 13 bar; hold time, 60 s)
Particle size range (mm)

106152
600850

Total volatile yield (%, daf)

He

Steam (15%),
CO2 (18%), N2
(67%)

51.0
44.7

71.7
57.8

carbon and some attention has been paid to the consequences of this particle size coal as feed.
It would have been desirable to conduct large-particle
bench-scale experiments in the wire-mesh reactor.
However, in this instrument, heat transfer at higher temperatures takes place mostly by radiation from the meshi.e.
from directly above and below the particle. For particles as
large as 600850 mm, the probability of uneven heating
would be much higher than in the case of smaller particles;
this would have introduced a spurious parameter. It was
therefore decided to perform these tests in the fluidised
bed reactor, where the major disadvantage was tar loss
due to cracking within the bed and in the reactor freeboard
[6,16].
Table 6 compares sample weight loss during the pyrolysis
and gasification of Daw Mill coal particles of two distinct
particle size ranges: 106152 and 600850 mm. The
samples for both size ranges were crushed from a single
large lump of coal. As before, the pyrolysis experiments
have been conducted in He; however, a mixture of steam/
CO2/N2 was used in the gasification experiments, to simulate the composition of gas at the exit from the spout of the
pilot scale gasifier. The temperature and pressure of this
limited set of experiments were chosen to reflect pilotplant conditions as closely as possible: 13 bar and 950C.
As an aside, the pyrolysis weight loss observed for the
106152 mm fraction suggests that this single lump of coal
was composed of material which was more reactive
compared to the already ground standard laboratory sample.
However, to the extent that the small particle sample was
prepared from further crushing of the large particles
themselves crushed and sieved from the original lump of
coal, the values given in Table 6 are internally consistent.
The pyrolysis data in Table 6 indicate that increasing the
particle size gives rise to a volatile loss of the order of 6%. It
is likely that this is due to re-deposition and re-polymerisation of part of the tar precursors, due to the longer path that

800

Y. Zhuo et al. / Fuel 79 (2000) 793802

Table 7
Combustion reactivities of chars from fluidised bed experiments carried out
to investigate the effect of particle size (coal: Daw Mill (washed); temperature: 950C; pressure: 13 bar; hold time: 60 s)

106152 mm coal
He
Steam/CO2/N2
600850 mm coal
He
Steam/CO2/N2
a

Rmax (% min 1, daf)

(t1/2) a (min)

0.42
0.28

150.8
228.8

0.30
0.24

222.9
314.5

Time required for 50% burnout.

has to be traversed. One consequence of such tar re-deposition is expected to be the formation of pyrolysis derived
secondary char within the pores, leading to reduced char
reactivity. In gasification, the difference between conversions of the different size particles were even greater
confirming the delaying effect due to increased particle
diameters.
However, for the 106152 mm range, the extents of
weight loss were lower, compared with results from the
same reactor (or the wire-mesh instrument) when using
pure CO2 or steam (cf. Figs. 1 and 3). It must be concluded
that the survival of the unreactive char in the pilot-scale
reactor must have also been related to the relatively low
reactivity of the gaseous medium.
Table 7 presents combustion reactivities of chars from
this set of experiments. The results for the 106152 mm
fraction show an already observed trend, where pyrolysis
chars appear somewhat more reactive than the gasification
chars; differences were admittedly small, but the 50% burnout times (t1/2) are sufficiently well spread out to confer
experimental significance to these observed differences.
Similarly chars from the larger particles were found to be
only slightly less reactive than chars prepared from the 106
152 fraction; once again the corresponding t1/2 values
suggest these differences to be experimentally significant.
All char reactivity values in this set were small. The
few available data points in Table 5 suggest that
combustion reactivities of chars from 60-s steam gasification experiments were also quite low. We have tended
to understand those low reactivities in terms of the high
conversions in steamthe residues of spent carbon
retaining few active centrescoupled with the time at
temperature already having been shown to reduce the
char reactivity quite significantly. However, in the
case of the larger particle chars (Table 7), conversions
in steam were clearly and very significantly loweryet
their reactivities were also lower.
It appears likely that the additional lower reactivity of
the large particle chars was caused by the lower activity
induced by the retained and re-polymerised tar precursors.
However, differences were small and the data cannot be

considered as conclusive in this respect. Clearly, the effect

of residence time at 1000C was more important in determining the reactivity of these chars.
3.4. Implications for commercial scale low-temperature
fluidised-bed gasifiers
Whilst particle heating rates in fluidised bed gasifiers are
estimated to be in excess of 1000 K s 1, such high heating
rates would only apply to external boundaries of the particles. Howard and Anthony [17] have calculated that, when
heated at 1000 K s 1, the temperature distribution within
coal particles of diameter greater than 100 mm diameter
could not be considered to follow the temperature at the
boundary. The rate at which the temperature distribution
is propagated within larger particles is governed primarily
by the thermal conductivity of the mass within the particle.
This gives rise to slower heating rates within the mass of the
particle. Data from the fixed-bed (hot-rod) reactor was
useful to the extent that it simulated coal behaviour in a
slow heating regime.
When large particles are fed to a low-temperature gasifier (9501000C), the sequence of processes taking place
may be summarised as follows: due to pyrolytic devolatilisation (pushing outwards), the gasifying medium experiences a delay in reaching external surfaces of the particle.
Gasification would then take place by a combination of a
shrinking core mechanism and by intraparticle diffusion of
the reactant gas. Of these two, the predominant mechanism
by which the particle is consumed depends on the magnitude of intraparticle diffusion resistances. However, the time
taken for intraparticle diffusion, as well as the time taken for
the core to shrink, increases rapidly with increasing particle diameter.
Within this framework, it is clear from our data that
unless the particle is consumed quickly, the mass of char
deactivates to very low levels of reactivity, e.g. within about
10 s of exposure to temperatures in the region of 950
1000C. For particles within the 600850 mm diameter
range, the time it takes for the gasification reaction to
reach completion is clearly too long, and the char rapidly
deactivatesapparently by an annealing process.
Further, conversions of small (106152 mm) particles in
experiments simulating the gas composition at the exit of
the spout in the pilot-scale gasifier were lower, compared to
experiments using pure steam (or even pure CO2). Thus the
survival of unreactive char in the pilot-scale reactor must
have been also related to the relatively low reactivity of the
ambient gaseous medium. The long time elapsed during the
reaction between the large particles and the low reactivity
gaseous medium appears to have led to the observed low
char reactivities.
For pilot or full-scale gasifier design, the challenge, therefore, is to formulate alternative fluidised bed gasifier operating strategies, which allow the use of smaller particle
sizes. Increasing the reactivity of the gas medium is less

Y. Zhuo et al. / Fuel 79 (2000) 793802

801

4. Summary and conclusions

swamped by deactivation due to time-at-temperature.

4. In low temperature gasifiers, the combined effect of
longer times required for consuming larger char
particles and the rapid deactivation of chars by
(10 s) exposure to gasification temperatures tends
to lead to appreciable levels of residual char removal.
5. Mathematical models form an important part of the
design process and are used as a predictive tool for
assessing commercial prospects. However, detailed
knowledge of rates of competing gasification reactions is required when setting up mathematical
models of the process. Findings outlined above indicate that coal/char reactivities may decrease to less
than one third of its zero second value during the
first 10 s at peak temperatureunderlining some
the difficulties of establishing reactor simulations,
which include fundamental kinetic models of coal
gasification.

Comparison with bench-scale experiments has been used

to evaluate strategies by which fuel conversion could be
improved in pilot-scale low-temperature (9501000C)
gasifiers.

The kinetics of char deactivation as a function of time-attemperature would appear to play an important role in determining eventual gasification conversions as well as the gasification kinetics itself.

straightforward. In the pilot plant, the CO2 present in the

system is formed as the product of combustion reactions
supplying the process heat. The air/steam ratio was used
as a means of controlling the temperature in the combustion
zone (spout) to below the limit set by ash sintering. Consequently, the steam content of the feed gas may not be varied
independently, to take advantage of its greater reactivity.
However, changing steam rates may act as a means of maintaining the gas quality if a lower temperature is needed, as
this would be achieved by raising the steam and lowering
the air input. Clearly, attempts to improve gasification
conversions in the pilot-scale reactor by reducing particle
size and also possibly by increasing the reactivity of the gas
stream would require arriving at a new set of optimum
operating conditions.

1. Above 10 bar, conversions and extents of gasification

determined in bench-scale experiments carried out with
small (106152 mm) particles were significantly higher
than total conversions and estimated steam gasification
levels in the pilot-plant gasifier. The data clearly showed
that lower overall conversions observed in the pilot scale
gasifier do not correspond to levels of reactivity that may
be expected from the properties of the original feedstock
itself.
2. During pyrolysis in helium, particle size was found to
affect sample weight loss significantly. In gasification,
the effect was more pronounced, extending in one particular case to differences of the order of 1518%, for an
experiment of 60 s duration. Conversions also depended
heavily on the reactivity of the gaseous medium: conversions in CO2 were systematically lower than those in
steam; conversions in both gases increased monotonically with pressure.
3. Combustion reactivities of residual chars diminished
rapidly with time-at-temperature. At 1000C, char reactivities declined sharply between 0 and 10 s holding time,
whilst the effect of increasing pressure was experimentally significant but much smaller. Tar retention in larger
particles has been observed to adversely affect sample
weight loss in pyrolysis and probably also in gasification.
However, the effect of the secondary char in reducing
char reactivity appears to be small, compared to the effect
of deactivation due to long (10 s or longer) exposure
times at temperature 1000C.
Under otherwise similar reaction conditions, pyrolysis
chars were found to be more reactive than CO2-gasification chars; steam-gasification chars were the least reactive. However, above 10 s, the effect was rapidly

Acknowledgements
The authors would like to acknowledge the financial
support received from the European Union under Contract
Nos JOF3/CT95/0018 and ECSC 7220-ED/075, and from
the British Coal Utilisation Research Association
(BCURA)/Department of Trade and Industry under
Contract No. B38. The authors would also like to thank
Mitsui Babcock Energy Ltd (owners of the British Coal
gasifier technology) for permission to publish data from
the ABGC gasifier.
References
[1] Dawes SG, Cross PJ, Minchener AJ. UNECE Symposium, Helsinki,
Finland Development of the British Coal Topping Cycle. May 1993.
[2] Johnson JL. In: Elliot MA, editor. Fundamentals of coal gasification,
chemistry of coal utilisation, 1981 chap. 23.
[3] Guell AJ, Kandiyoti R. Energy and Fuels 1993;7:943.
[4] Gibbins JR, Kandiyoti R. Energy and Fuels 1989;3:670.
[5] Messenbock RC, Dugwell DR, Kandiyoti R. Energy and Fuels 1999,
in press.
[6] Megaritis A, Zhuo Y-Q, Messenbock R, Dugwell DR, Kandiyoti R.
Energy and Fuels 1998;12:144.
[7] Pindoria RV, Lim J-Y, Hawkes JE, Lazaro M-J, Herod AA, Kandiyoti
R. Fuel 1997;76:1013.
[8] Radovic LR, Walker PL, Jenkins RG. Fuel 1983;62:849.
[9] Megaritis A, Messenbock RC, Collot A-G, Zhuo Y, Dugwell DR,
Kandiyoti R. Fuel 1998;77:1411.
[10] Lim J-Y, Chatzakis IN, Megaritis A, Cai H-Y, Dugwell DR,
Kandiyoti R. Fuel 1997;76:1327.
[11] Gibbins JR, Gonenc ZS, Kandiyoti R. Fuel 1991;70:621.
[12] Howard JB. In: Elliot MA, editor. Chemistry of coal utilization,
Supplementary volume II, New York: Wiley, 1981. p. 665.

802

Y. Zhuo et al. / Fuel 79 (2000) 793802

[13] Guell AJ, Cai H-Y, Dugwell DR, Kandiyoti R. Fuel Proc Tech
1993;36:259.
[14] Gonenc ZS, Gibbins JR, Katheklakis IE, Kandiyoti R. Fuel
1990;69:383.

[15] Messenbock RC. PhD thesis, University of London, 1998.

[16] Stiles HN, Kandiyoti R. Fuel 1989;68:275.
[17] Howard JB, Anthony DB. AIChE J 1976;22:625.