Pressure drop prediction of a gasifier bed with cylindrical biomass pellets

Applied Energy 113 (2014) 258–266

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Pressure drop prediction of a gasifier bed with cylindrical biomass
pellets
Duleeka Sandamali Gunarathne ⇑, Jan Karol Chmielewski, Weihong Yang
Royal Institute of Technology, Department of Material Science and Engineering, Division of Energy and Furnace Technology, Brinellvägen 23, 100-44 Stockholm, Sweden

h i g h l i g h t s
 An equation was developed for pressure drop prediction with shrinking effect.
 Graphical representations of correlation constants were introduced.
 This would provide a guide to select pellet size and designing a grate.

a r t i c l e

i n f o

Article history:
Received 6 March 2013
Received in revised form 26 June 2013
Accepted 13 July 2013
Available online 8 August 2013
Keywords:
Biomass
Gasification
Fixed bed
Pressure drop

a b s t r a c t
Bed pressure drop is an import parameter related to operation and performance of fixed bed gasifiers. Up
to date, limited literature is found on pressure drop prediction of beds with cylindrical pellets and none
was found for gasifying beds with cylindrical pellets.
In this paper, an available pressure drop prediction correlation for turbulent flows in a bed with cylindrical pellets which has used equivalent tortuous passage method was extended for a gasifier bed with
shrinking cylindrical pellets and for any flow condition. Further, simplified graphical representations
introduced based on the developed correlation can be effectively used as a guide for selecting a suitable
pellet size and designing a grate so that it can be met the system requirements.
Results show that the method formulated in the present study gives pressure drop approximation
within 7% deviation compared to measured values with respect to performed runs. Available empirical
correlation with modified Ergun constants for cylindrical pellets gave pressure drop within 20% deviation
after the effect of shrinkage was taken into account.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction
Biomass gasification is a promising renewable energy technology for supplying thermal energy and generating electric power.
Nowadays, pelletized biomass is widely used in order to overcome
some problems when using conventional biomass in thermal
applications like gasification including logistic problems due to
low bulk density, non-uniformity of fuel, low energy density, etc.
Pressure drop is an important factor in fixed bed gasification of
Biomass. Most common and widely used method for predicting
pressure drop in a packed bed is using Ergun equation which has
viscous and inertial terms corresponding to laminar and turbulent
flow conditions. One limitation of this model when applying for a
gasifier bed with biomass pellets is due to the particle shape which
is essentially cylindrical shape when considering pellets. Limited
literature is available in pressure drop prediction for a bed with
cylindrical pellets.

⇑ Corresponding author. Tel.: +46 8 790 8402; fax: +46 8 207 681.
E-mail address: (D.S. Gunarathne).
0306-2619/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
/>
One research group [1] has considered this effect and has developed an equation for pressure drop in packed bed with cylindrical
shaped particles by using equivalent tortuous passage method. But,
the equation is limited to turbulent flow conditions and not valid
for a bed with laminar or transition flow conditions.
Some investigators [2] have developed an empirical correlation
for Ergun constants for a bed of cylindrical particles by referring
the sphericity of particles. But, this correlation does not show
strong validity due to scatter of data and suitable only for a rough
approximation of the pressure drop. Lack of theoretical background is another limitation for applying this correlation.
Further considering these models, none of the above models for
cylindrical particles are developed for active beds of particles. In a
gasifier, particles participate in the reaction and therefore particle
size and the porosity of bed varies with time and along the height
of the bed. Even if steady state condition is considered, the spacial
variation of porosity has to be taken into account.
Some researchers [3] have addressed this issue on a downdraft
gasifier but with particles in spherical shape and hence Ergun equation and its’ another variation called Macdonald correlation have


259

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266

Nomenclature
A
a

B
c
c0
D
De
d
d0
dp
ER
K1
K2
K3
Kt
L
LHV
l
mbiomass

modified Ergun constant for viscous term (–)
wetted perimeter (m)
modified Ergun constant for inertial term (–)
length of cylindrical particle at any time (m)
initial length of cylindrical particle (m)
diameter of gasifier (m)
equivalent diameter of tortuous passage (m)
diameter of cylindrical particle at any time (m)
initial diameter of cylindrical particle (m)
equivalent particle diameter (m)
equivalence ratio (–)
constant depend on c0, d0 and x (for inertial term) (–)

constant depend on c0, d0 and x (for viscous term) (–)
constant depend on c0, d0 and x (for Rep) (–)
constant depend on roughness of particle and packing
tortuosity (–)
length (height) of the bed (m)
lower heating value of biomass (MJ/kg)
equivalent length of tortuous passage (m)
mass of single biomass particle (kg)

been used along with considering the wall effects. Another group [4]
has focused on cylindrical wood particle in a fluidized bed considering shrinking effect but only pyrolysis conditions. An interesting literature [5] was found for a coal gasifier and they have found
pressure drop variations within different zones in the gasifier by
using Ergun equation for each zone separately. This method has
incorporated lots of experimental data and may be successfully
used for that specific commercial gasifier model but not for any type
of gasifier. Therefore, none of the above cases can be used for predicting pressure drop in any fixed bed gasifiers of cylindrical pellets.
One of our previous works [6] concentrated on developing a
model for prediction of pressure drop due to grate-bed resistance
of a gasifier. As the second step of that, with the objective of filling
the gap on pressure drop prediction of gasifier beds with cylindrical pellets, here we focus on the bed resistance. Considering limitations of previous models, an equation is developed based on
the model predicted in the literature [1] including the effect of
laminar and transition flow conditions and also the effect of
shrinkage of particles during gasification and will be verified based
on experimental data. Further, it will also compare with the empirical correlation available for cylindrical pellets [2] which will also
be upgraded by taking shrinking effect into account.

mchar
DP
Rep
rH

Sp
T
u
Vp
v
x

mass of single char particle (kg)
pressure drop (Pa)
particle Reynolds number (–)
hydraulic radius (m)
particle surface area (m2)
temperature (°C)
superficial velocity (m/s)
particle volume (m3)
velocity of flow through tortuous passage (m/s)
mass conversion (–)

Greek letters
e
porosity (–)
U
sphericity (–)
k
angle of inclination of tortuous passage to the mean
flow (°)
l
viscosity (Ns/m2)
q
density (kg/m3)

This unit is incorporated with feed gas preheater, updraft gasifier,
fuel feeding system and producer gas post combustion unit. Detailed description of this experimental facility is available elsewhere [7].
Biomass pellets stored in the feed tank is transported to the gasifier via screw conveyor. The frequency of feeder is correlated with
the feeding rate. Required frequency set point is predetermined in
order to achieve specific biomass feed rate.
Preheated air from the preheater is introduced to the gasifier at
the side of bottom section below the grate. The system has the
facility to add steam to the feed stream if required. The flow of
hot gases and biomass is countercurrent. The grate facilitates to
build up a fixed bed of biomass and small particles left after considerable reaction can pass through the grate and collected below.
The producer gas which is flown upwards, leave the gasifier at the
side of the top section and is burned out at the combustion
chamber.
2.3. Experimental procedure and data reduction
The feeder was pre-calibrated with biomass pellets used in the
experiment. Once the air temperature was reached around 1000 °C

2. Materials and methods
2.1. Materials
The biomass pellets used in the experiment (Fig. 1) were supplied by Boson Energy S.A.
The length distribution of pellets considering 50 numbers of
pellets is given in Fig. 2.
It can be seen that the majority of pellets are in the range of 11–
15 mm in length. Physical properties of pellets (before the experiment) can be summarized as in Table 1.
Pellet proximate and ultimate analysis along with the heating
value is given in Table 2.
2.2. Gasification system
Gasification experiments were carried out in updraft High Temperature Agent Gasifier (HTAG) unit with 0.4 m diameter (Fig. 3).

Fig. 1. Biomass pellets.


260

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266

Fig. 2. Length distribution of pellets.

Table 1
Physical properties of pellets.
Parameter

Value

Diameter (mm)
Mean length (mm)
Moisture content (%)
Bulk density (kg/m3)
Particle density (g/cm3)
Porosity – e
Sphericity – U
Diameter of equivalent spherical particle (mm) – dp

8
14
9.8
603
1.09
0.44

0.85
11

Table 2
Composition of pellets.
Parameter

Value

Moisture content at 105 °C
Ash cont. at 550 °C
LHV
Volatile matter
Bulk density

9.8%
1.9% (dry)
16.6 MJ/kg (as received)
81.2% (dry)
603 kg/m3

Ultimate analysis
Carbon C
Hydrogen H
Nitrogen N
Oxygen O

49.4% (dry)
5.9% (dry)
0.17% (dry)

42.6% (dry)

by using preheater, desired feed rate was achieved by adjusting
frequency of feeder.
Temperatures inside the gasifier were measured with type S
thermocouples located along the reactor height and recorded by

data acquisition system connected to a PC. Pressure inside the
gasifier below the grate and three more points above the grate
were measured with digital manometers so that bed pressure drop
can be calculated. It was assumed that the horizontal gradient of
temperature and pressure is not significant.
Syngas composition was measured with Gas chromatography
(GC). Tar samples were collected from the gas outlet pipe and analyzed later for quantity and composition.
For each run, 20 min time interval was selected for analysis.
This time interval was selected based on stable temperatures and
gas compositions. Average values of gas compositions and temperatures within this time interval were taken for analysis.
From ultimate analysis of fuel, the average chemical formula of
pellets was obtained as CH1.43O0.65 and it was used to calculate the
stoichiometric air to fuel ratio for calculating Equivalence Ratio
(ER).
Gas flow rate was calculated by applying Nitrogen balance over
the gasifier.
The average gas properties such as density and viscosity within
the gasifier bed were calculated by taking volume average with gas
composition data at average bed temperature.
3. Results and discussion
3.1. Process performance
Table 3 gives experimental data and Figs. 4–6 show variation of
temperature along the gasifier height, gas compositions and characteristic ratios respectively. Table 4 gives the tar composition and

tar characteristic ratios of each run.
From temperature data, we see different bed temperature
behaviours with two cases. Run 1 with low ER shows gradual temperature drop throughout the bed height and run 2 with high ER
shows high temperature adjacent to the grate with sudden drop
after that. Then it can be expected with run 2, CO2 and H2O generated by exothermic combustion reactions at high temperature
zone near the grate has reduced to CO and H2 by endothermic boudouard and water gas reactions at the subsequent low temperature
region. As a result, high CO and H2 content can be seen with run 2.
They were respectively 4% and 5% increment compared to run 1.
In overall, low temperature was seen throughout the gasifier
with low ER in run 1 and comparatively high hydrocarbon and also
tar content was observed compared to run 2 due to cracking reactions. Even with low CO and H2 contents, as a result of high CH4
and CxHy contents which were around 11% and 140% higher than
run 2, LHV is slightly higher with run 1. However, high CO and
H2 content with run 2 resulted in high gas yield and hence considerably higher efficiency.

Fig. 3. HTAG system.


261

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266
Table 3
Experimental data.
Run

1
2
a

Table 4

Tar composition data.

Dry
biomass
(kg/h)

Feed gasa
(N m3/h)

ER

54.12
45.1

69
69

0.22
0.27

Quantity lg/100 ml
Run 1
(ER = 0.22)

Run 2
(ER = 0.27)

Benzene
Toluene
m/p-Xylene

o-Xylene
Indan
Indene
Naphthalene
2-Methylnaphthalene
1-Methylnaphthalene
Biphenyl
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluorantene
Pyrene
Phenol
o-Cresol
m-Cresol
p-Cresol
Unknown
Total (g/N m3)

269.7
112.7
11.5
7.4
4.1
118.8
219.6
22.8
31.5

8.8
97
3
21.5
25.8
7.6
10.3
0
45.6
1.9
7.1
1.4
223
12.5

186.5
43
3.1
9.8
4.7
22.4
55.7
0
16.1
0
19.4
0.6
2.8
0
2.2

0
0
4.1
0
0
0
69.1
4.4

Characteristic ratios
C2H6/(C2H4 + C2H2)
Phenols/aromatics
Indene/naphthalene

0.29
0.1
0.54

0.1
0.03
0.4

LHV
(MJ/
N m3)

Gas yield
(N m3/kg dry
biomass)

Efficiency
(%)

Component

5.59
5.55

2.18
2.63

66.42
79.4

Feed gas contains 17% O2, 81% N2 and 2% CO2.

Fig. 4. Temperature profile along the gasifier height.

Fig. 5. Gas compositions.

high temperature and longer bed. There was no significant difference seen with H2/CO ratio of two cases.
Significant reduction of almost all the tar components was seen
with run 2. Tar characteristic ratios were also reduced and it represents that high temperature and longer bed has a positive impact
on tar decomposition reactions.
Referring to Table 5, it was observed a considerable bed height
achieved with each run. When bed height is higher, residence time
for both solid and gas phase reactions are larger and it is reflected
by high CO and H2 content, gas yield and gasification efficiency obtained with run 2. Specially, significant reduction of tar content is
also positive.
However, the drawback of such large bed is large pressure drop

of the system which ultimately affects the system performance.
Therefore, prediction of pressure drop of a gasifier bed is a quite
interesting topic for anyone concerning the system performance.
3.2. Prediction of pressure drop
3.2.1. Developing the correlation
Total pressure drop through a gasifier bed is mainly a sum of
pressure drop through the particle bed and pressure drop through
the grate. However, in this study, grate resistance can be considered as negligible since grate opening area is high as much as
40% and the grate thickness is low which is 6 mm.
Literature [1] has derived Eq. (1) based on equivalent tortuous
passage method for pressure drop DP over a bed height of L in a
turbulent flow using Blasius smooth pipe equation for a packed
bed with cylindrical particles of diameter d and length c,

Fig. 6. Gas characteristic ratios.

When characteristic ratios are considered, CO/CO2 ratio was
higher with run 2 CH4/H2 and CxHy/CH4 ratios were higher with
run 1. CH4/H2 and CxHy/CH4 ratios show the effectiveness of hydrocarbon cracking and cracking shows more effective in run 2 due to

Table 5
Bed heights and pressure drop.
Run

Bed height (m)

Pressure drop across the bed (Pa)

1
2

0.55
0.6

1000
1190


262

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266



ð1 À eÞ5=4

DP ¼ LK t qu2

1 1
þ
2c d

e

3

5=4 

1=4

l
qu

Kt is a constant combining roughness of the particles and packing
tortuosity all together and has determined experimentally and related to porosity e as follows.

K t ¼ 112e3:2
Since they have considered only the inertial term of pressure
drop, it can be modified to fit to laminar or transitional flows also
by adding a viscous term.
Hagen–Poiseuille equation for pressure drop in laminar flow is,

DP ¼

32lv l
D2e

Fixing to the definitions in [1] which are tabulated in Table 6
assuming equivalent inclined passage with an angle k to the direction of mean flow, Hagen–Poiseuille equation can be re-arranged
as,

DP ¼ 32luL

ð1 À eÞ2



2
1 1
þ

2c d

e3

ð2Þ

Then, by combining Eqs. (1) and (2) and rearranging, Rangel
equation can be modified for any flow condition as,

 

DP
1
K1
K2
¼ 1=4 þ
L
qu2
Rep
Rep

ð3Þ

qu

where Rep ¼

ðlÞ
K3

K 1 ¼ 112e0:2 ð1 À eÞ

K 2 ¼ 32

ð1 À eÞ



1 1
þ
2c d



e

3

1 1
þ
2c d







1 1
K 3 ¼ ð1 À eÞ

þ
2c d

U¼

"
#1=3
2:25 Ã ðdc Þ2
À Á3
0:5 þ dc

ð5Þ

Then, porosity e of a bed with cylindrical particles is obtained as
a function of particle size c and d.
3.2.1.2. Shrinking effect of particles. Due to the reaction happening
in the gasifier bed, the particle size is changing along the bed. This
results in change of sphericity and consequently the porosity of the
bed. Wall effect and thickness effects on porosity variation can be
neglected for the cases with tube to particle diameter ratio D/dp
and bed height to particle diameter ratio L/dp are high. According
to [9] the values should be D/dp P 10 and L/dp > 3 in order to neglect those effects. This assumption was applied here assuming
the gasification is done in a pilot scale unit with considerable
diameter compared to particle size and achieving considerable
height of bed which only necessitates pressure drop prediction.
Particle size of a reacting bed can be calculated by applying
mass balance for one particle and mass of char particle mchar and
mass of initial biomass particle mbiomass can be related as,

mchar ¼ ð1 À xÞmbiomass

Practically two types of size reductions can be expected in a
gasifier; fragmentation and conversion. Fragmentation can be taken as less important when it comes to wood pellets compared
to wood chips gasification due to high density of pellets [10].
Therefore, surface conversion was assumed to dominate in this
case. With surface conversion, density of biomass particle can be
taken as constant throughout the conversion period. Then, volume
of char particle and volume of initial biomass particle can be related same as above.
If initial length and diameter was considered as c0 and d0 it
becomes,
2

2

cd ¼ ð1 À xÞc0 d0

Graphical representations of above correlation constants for
typical biomass pellet sizes available in the market are annexed.
3.2.1.1. Relation of porosity and sphericity in a bed of cylindrical
particles. Some researchers [8] have formulated a relationship between porosity and sphericity U for loose random packing of cylindrical particles as given in Eq. (4). This correlation shows very good
agreement with their experimental data.

ln e ¼ U5:58 exp½5:89ð1 À Uފ ln 0:4

ð4Þ

The sphericity of a cylindrical particle depends on its length.
Very long or very short particles give low sphericity. The sphericity
of cylindrical particle is given by,
1=3

ð36pV 2p Þ
U¼
Sp

where Vp and Sp are cylinder volume and area respectively.
Table 6
Defining parameters in tortuous passage.
Parameter related to tortuous passage

Definition fixing to [1]

Velocity of flow
Equivalent length

u
v ¼ e cos
k

Equivalent diameter
Hydraulic radius
Wetted perimeter

Substituting for volume and area,

ð1Þ

l ¼ cosL k
De = 4rH
e
r H ¼ a cos

k
S

a ¼ Vpp ð1 À eÞ

ð6Þ

For a cylindrical wood pellet, assuming uniform thickness h is
reduced for a certain time period from all its dimensions [11], after
a certain time new length c and diameter d of particle is given by,
c = c0 À 2h and d = d0 À 2h
Avoiding unknown h,

c À d ¼ c 0 À d0

ð7Þ
c
d

Knowing c0, d0 and x, can be obtained from Eqs. (6) and (7) and
used in Eq. (5), in order to calculate sphericity. And then, sphericity
can be used in Eq. (4) for calculating porosity. These values along
with flow properties such as velocity, density and viscosity can
be used in Eq. (3) in order to calculate pressure drop along the gasifier bed for a known conversion and bed height.
3.2.2. Calculation of pressure drop
Conversion x at the top of the bed is 0 and at the bottom x is assumed to be 1. The average mass conversion within the bed can be
calculated based on the C, H and O molar balance Table 7 summarizes the molar inputs, outputs and also accumulated in char.
From molar rates of each species accumulated in char which is
equal to difference in input and output, hourly char generation can
be calculated and it is 12.78 kg/h and 12.68 kg/h respectively in

two cases. Then, average mass conversion x in the bed is 0.79
and 0.75 respectively.
With conversion values calculated, referring to Section 3.2.1, c,
d, U and e can be calculated and given in Table 8.
The particle diameter and average length has reduced respectively from 8 mm and 14 mm initial values to around 4.5 mm
and 10 mm at the average conversion. With reduced particle sizes


263

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266
Table 7
C, H, O molar balance.
Run

Description

C (kmol/h)

H (kmol/h)

O (kmol/h)

1

Input (biomass & feed gas)
Output in syngas
Char

2.29

1.94
0.35

3.85
2.47
1.38

2.93
2.48
0.45

2

Input (biomass & feed gas)
Output in syngas
Char

1.92
1.19
0.73

3.21
2.49
0.72

2.64
2.44
0.20

gas compositions. Then, these values along with c, d and e can be

used as inputs to the Eq. (3). Table 9 represents all the parameters
and calculated bed pressure drop for both runs.

3.2.3. Incorporating shrinking effect into available empirical
correlation for comparison
For cylindrical particles some researchers [2] have obtained a
relationship with the sphericity and Ergun constants A and B as given in the following equation:

"

DP ¼ L A
Table 8
Pellet properties after conversion.
Initial length
range (mm)

Run 1
c

d

U

e

Run 2
c

d

U

e

1–5
6–10
11–15
16–20
21–25
26–30
Weighted
average

1.09
4.76
9.33
14.14
19.03
23.96
9.7

6.09
4.76
4.33
4.14
4.03
3.96
4.4

0.613

0.874
0.823
0.759
0.706
0.664
0.817

0.558
0.404
0.416
0.443
0.476
0.509
0.421

1.23
5.04
9.65
14.46
19.36
24.29
10

6.23
5.04
4.65
4.46
4.36
4.29
4.7

0.637
0.874
0.828
0.767
0.717
0.676
0.822

0.534
0.404
0.415
0.439
0.469
0.499
0.419

ð1 À eÞ2 l

e3 U2 d2p

uþB

ð1 À eÞq 2
u
e3 Udp

#
ð8Þ

1:75
where A ¼ U150
3=2 and B ¼
U4=3
By incorporating shrinking effect this equation can be improved
for a reacting bed. To do this, sphericity and modified Ergun constants were calculated for each initial length interval and their
average values are given in Table 10.
The modified constants calculated for two cases are as follows.
For Run 1,

"

DP ¼ L 204

ð1 À eÞ2 l

e3 U2 d2p

u þ 2:3

ð1 À eÞq 2
u
e3 Udp

#

For Run 2,

"

DP ¼ L 202

Table 9
Summary of parameters for bed pressure drop calculation.
Parameter

Run 1

Run 2

T
u

887
1.11
0.2692
4⁄10À5
9.7
4.4
0.421
7.03
56.77
0.55
1005

907
1.13
0.2662
4.02⁄10À5
10

4.7
0.419
6.92
60.53
0.6
1274

q
l
c
d

e
Kt
Rep
L
DP

Table 10
Summary of calculating modified Ergun constants A and B.
Initial length
range (mm)

Run 1
Average
sphericity

A

B

Run 2
Average
sphericity

A

B

1–5
6–10
11–15
16–20
21–25
26–30
Weighted average

0.613
0.874
0.823
0.759
0.706
0.664
0.817

313
184
201
227
253

277
204

3.36
2.09
2.27
2.53
2.78
3.02
2.30

0.637
0.874
0.828
0.767
0.717
0.676
0.822

295
184
199
223
247
270
202

3.19
2.09
2.25

2.49
2.73
2.95
2.28

it can be expect that low porosity since small sized particles pack
more tightly than large size ones. Proving this, the initial porosity
0.445 has reduced up to 0.42 at achieved conversion.
With conversion, pellets get small and porosity is reduced.
Therefore, porosity at the top of the reacting bed is highest and it
is lowest at the bottom.
Density, viscosity and superficial velocity of gas flow inside the
bed can be approximated by bed temperature, gas flow rate and

ð1 À eÞ2 l

e3 U2 d2p

ð1 À eÞq 2
u þ 2:28 3
u
e Udp

#

Ergun indices obtained are 35% and 31% increased respectively
compared to original Ergun constants which are 150 and 1.75 for
viscous and inertial terms respectively. When bed is composed of
cylindrical particles, the pressure drop is higher compared to packing spherical particles. The orientation of particles, tortuosity and
wetted surface are blamed regarding this increase [2].

3.2.4. Validation with experimental data
Pressure drop results calculated with developed correlation and
empirical correlation can be compared with experimental data as
given in Fig. 7.
The method formulated in the present study gives better
approximation with only 7% maximum error with respect to performed two runs. The available empirical equation was able to predict the pressure drop within 20% interval after shrinking effect
was taken into account.

Fig. 7. Comparison of pressure drop results.


264

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266

Fig. A1. Variation of K1 with conversion for pellets of 8 mm diameter.

Fig. A2. Variation of K2 with conversion for pellets of 8 mm diameter.

Fig. A3. Variation of K3 with conversion for pellets of 8 mm diameter.


D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266

Fig. A4. Variation of K1 with conversion for pellets of 6 mm diameter.

Fig. A5. Variation of K2 with conversion for pellets of 6 mm diameter.

Fig. A6. Variation of K3 with conversion for pellets of 6 mm diameter.

265


266

D.S. Gunarathne et al. / Applied Energy 113 (2014) 258–266

4. Conclusions
A correlation for pressure drop prediction in a gasifier bed with
cylindrical particles was proposed, compared with available empirical correlation for cylindrical pellets and verified with experimental data.
Based on the developed correlation, simplified graphical representations were introduced for commonly available pellet sizes in
order to reduce the calculation effort. The plots developed can be
effectively used as a guide for selecting suitable pellet size and
designing a grate so that it can be met the system requirements.
Acknowledgements
Authors like to acknowledge KIC-innoenergy project which provided the financial support and Boson Energy S.A. which provided
the biomass samples for experimental work.
One of authors, Duleeka Sandamali Gunarathne would like to
acknowledge the financial supporting from the European Commission. This publication reflects the views only of the author, and the
Commission cannot be held responsible for any use which may be
made of the information contained therein.
Appendix A. Graphical representation of correlation constants
It was reported that pellet size has the more impact on the
shrinking behavior, not the composition of pellet [12,13]. Commercially available pellets are commonly found with 6 mm and 8 mm
in diameter with maximum length to diameter ratio being 5 [14].
Then, for those pellets, following figures can be used to find the
K values to be used in the Eq. (3) at any conversion if the initial particle size distribution is known.
According to Figs. A1–A6, very rapid increase of K values and
hence the pressure drop can be seen at the end of the conversion

period which is happening in the bottom of the bed. By having a
grate opening area large enough to maintain conversion below
0.9 may be beneficial in this case depending on the ability of the
system to overcome the pressure drop. Therefore, someone can

use these figures as a guide for designing a suitable grate for the
system. On the other hand, smaller the pellet size, larger the pressure drop in the system and it is also clearly seen in these figures.
With lower length to diameter ratio and small diameter, all the K
values and hence the pressure drop will be high. Therefore, this
can be another guide for selecting a suitable pellet size for the system requirements.
References
[1] Rangel N, Santos A, Pinho C. Pressure drop in packed shallow beds of
cylindrical coke stoppers. Trans IChemE 2001;79(Part A).
[2] Nemec D, Levec J. Flow through packed bed reactors: 1. Single-phase flow.
Chem Eng Sci 2005;60:6947–57.
[3] Sharma AK. Modeling fluid and heat transport in the reactive, porous bed of
downdraft (biomass) gasifier. Int J Heat Fluid Flow 2007;28:1518–30.
[4] Sreekanth M, Kolar AK. Progress of conversion in a shrinking wet cylindrical
wood particle pyrolysing in a hot fluidized bed. J Anal Appl Pyrolysis
2009;84:53–67.
[5] Luckos A, Bunt JR. Pressure-drop predictions in a fixed-bed coal gasifier. Fuel
2011;90:917–21.
[6] Donaj P, Izadpanah MR, Yang W, Blasiak W. Effect of pressure drop due to grate
and bed resistance on the performance of a downdraft gasifier. Energy Fuels
2011;25(11):5366–77.
[7] Lucas C. High temperature gasification of biomass in an updraft fixed bed batch
type gasifier. PhD thesis, ISBN 91-7178-067-X.
[8] Zou RP, Yu AB. Evaluation of the packing characteristics of mono-sized nonspherical particles. Powder Technol 1996;88:71–9.
[9] Foumeny EA, Roshani S. Mean voidage of packed beds of cylindrical particles.
Chem Eng Sci 1991;46(9):2363–4.

[10] Teixeira G, Van de Steene L, Martin E, Gelix F, Salvador S. Gasification of char
from wood pellets and from wood chips: textural properties and
thermochemical conversion along a continuous fixed bed. Fuel
2012;102:514–24.
[11] Nffiez C, Cruells M, Garcia-Soto L. A general shrinkage particle model for the
chemical dissolution of all types of cylinders and discs. Hydrometallurgy
1994;36:285–94.
[12] Erlich C, Björnbom E, Bolado D, Giner M, Fransson TH. Pyrolysis and
gasification of pellets from sugar cane bagasse and wood. Fuel
2006;85:1535–40.
[13] Erlich C, Fransson TH. Downdraft gasification of pellets made of wood, palmoil residues respective bagasse: experimental study. Appl Energy
2011;88:899–908.
[14] Cocchi M, Nikolaisen L, Junginger M, Goh CS, Heinimö J, Bradley D, et al. Global
wood pellet industry market and trade study, IEA Bioenergy Task 40:
Sustainable international bioenergy trade; 2011.