Professional Documents
Culture Documents
1 s2.0 S0360544220307313 Main
1 s2.0 S0360544220307313 Main
Energy
journal homepage: www.elsevier.com/locate/energy
a r t i c l e i n f o a b s t r a c t
Article history: The effects of homogenous and heterogeneous moisture on wet wood pyrolysis are investigated using
Received 10 June 2019 modified FireFOAM. The wet wood pyrolysis model was used by considering one-step moisture drying
Received in revised form and one-step wood pyrolysis, which simplified the kinetic of intermediate solids. In homogeneous cases,
1 March 2020
the maximum mass fluxes of vapor and pyrolysate increase with the increased initial moisture content
Accepted 13 April 2020
Available online 21 April 2020
and there is only one peak on the mass flux curve of pyrolysate in the cases of wood thickness of 2 mm
and 5 mm, while two peaks appear in the cases of 12 mm and 20 mm. In heterogeneous cases, the
heterogeneous moisture significantly changes the mass fluxes and duration time of pyrolysate and vapor.
Keywords:
Pyrolysis
The detailed comparisons of the duration of vapor and the maximum mass fluxes of vapor and pyrolysate
Wet wood were also presented between homogeneous and heterogeneous moisture. This work provides a more
FireFOAM comprehensive understanding for the wet wood pyrolysis.
Moisture gradient © 2020 Elsevier Ltd. All rights reserved.
https://doi.org/10.1016/j.energy.2020.117624
0360-5442/© 2020 Elsevier Ltd. All rights reserved.
2 H. Liu et al. / Energy 201 (2020) 117624
types of modelling tools are available for pyrolysis analysis from basically exists in a non-uniform state. However, there is no
simple heat and mass balance models to advanced numerical research on this topic. In order to fill the knowledge gap in the
models. Amongst the various modellings, Computational fluid dy- numerical simulation of the wood pyrolysis with heterogeneous
namics (CFD) simulations have been widely used in the study of moisture. In the current study, the effect of the moisture gradient
biomass pyrolysis in the past two decades [12]. Some researchers was first focused on the pyrolysis of wet birch wood (Betula). So, on
have focused on modeling the wood pyrolysis by CFD to further the one hand, the effect of different initial moisture content on the
explore the pyrolysis kinetics of wood. M€ atzing et al. [13] developed pyrolysis of wet wood under the same external radiant heat flux
a 1-D reactor cascade model and found that the results obtained by was analyzed. On the other hand, the effect of different moisture
CFD are in good agreement with the measured data. Zeng et al. [14] gradients on the pyrolysis of wet wood under the same external
developed a 2-D unsteady CFD single-particle model to simulate radiant heat flux was also analyzed to gain a more comprehensive
the solar pyrolysis characteristics of beech wood. Soria et al. [15] understanding of wet wood pyrolysis.
reported a 2-D single particle model based on the CFD platform
ANSYS FLUENT 14.0 to study the solar pyrolysis of beech wood
2. Mathematical model
pallets. Gonza lez et al. [16] investigated the biomass gasification
process in a fixed bed by CFD numerical analysis. The mathematical
The large eddy simulation (LES) based FireFOAM [26] solver is
model is a transient 2-D CFD model, which is extended to simulate
used to simulate the pyrolysis process of wet wood. FireFORM
the gasification process by expanding the chemical kinetic mech-
within the OpenFOAM toolbox is a free, open source CFD software
anism and adapting the stages of pyrolysis, oxidation, and reduc-
for modeling fire and solid pyrolysis, non-premixed combustion,
tion. Kwiatkowski et al. [17] reported a 3-D model of wood
etc [27]. The 1-D diffusion model with the Arrhenius type in Fire-
pyrolysis and gasification using ANSYS Fluent CFD package. The
FOAM is used in the present study, which is a common simplifi-
results show a good agreement with the experimental data.
cation to describe the thermal degradation of pyrolysis that
Wickramaarachchi et al. [18] investigated the drying and pyrolysis
effectively reduces the complexity and computation time [28].
process of a thermally thick single wood particle by an unsteady 3-
The pyrolysis process of wet wood is mainly divided into two
D model which suggests that the gas and solid phases inside the
parts. The one is the moisture drying process while the other one is
particle are not in thermal equilibrium. Go mez et al. [19] took
wood pyrolysis. The expressions are described as follows:
several sub-models from literature to simulate the combustion of
solid biomass in packed beds. These sub-models in commercial CFD Moisture/Vapor (1)
code involve the thermal conversion of solid fuels and its interac-
tion with the gas phase. Kumar et al. [20] developed a CFD model to
Wood / Char þ Pyrolysate (2)
study the gasification thermochemical processes of rubber wood. Li
et al. [21] assessed the devolatilization of biomass under high- The evaporation rate of moisture (u_ m ) and the mass loss rate of
temperature conditions by a CFD model which is based on a mul- wood (u_ w ) are calculated by nth-order Arrhenius formula:
tiscale EulerianeLagrangian solver previously developed in the
nm
framework of OpenFOAM. Blondeau et al. [22] reported a CFD rYm Ea;m
u_ m ¼ ðrYm Þ0 Am exp (3)
model to investigate the pyrolysis of particles in pulverized-fuel ðrYm Þ0 RT
boilers. The model accounts for internal heat conduction, internal
gaseous convection, moisture evaporation and particle shrinkage. nw
rYw Ea;w
These developed models have been proven to have good pre- u_ w ¼ ðrYw Þ0 Aw exp (4)
ðrYw Þ0 RT
dictive ability. However, wood may be exposed to wet surroundings
in an actual industrial scenario. In order to develop a model for
where subscripts m and w denote moisture and wood, respectively.
more effective prediction, the increase in the complexity of wood
r is the averaged density of wood, moisture and char, Y is the mass
pyrolysis model is required to consider the effect of moisture on
fraction of each component, and n is the reaction order. A and Ea are
wood pyrolysis [10]. However, nowadays, relatively little research
the pre-exponential factor and activation energy, respectively. R is
has been carried out. Ding et al. [23] developed a wet wood py-
the universal gas constant (8.314J mol1 K1) and T is the absolute
rolysis model to study the mass fluxes of water vapor and pyroly-
temperature. The subscript 0 denotes initial conditions before
sate under various external radiation heat flux. However, the effects
heating. The values of n, A and Ea are listed in Table 1. Then, the heat
of different initial moisture gradients on the pyrolysis behavior
flux is expressed as follows:
were not considered. Shen et al. [24] investigated the pyrolysis and
ignition of wood. The sample of wood was fixed in a horizontal
q’’m ¼ u_ m Hm (5)
orientation and heated by a cone calorimeter which supplying heat
flux from 0 to 80 kW m2. Then, a 1-D model was proposed to
investigate the effect of heat flux and moisture content on wet q’’w ¼ u_ w Hw (6)
wood pyrolysis. But some deficits were observed between the
calculated and experimental results. Pozzobon et al. [25] studied where Hm and Hw are the heat of moisture evaporation and heat of
the beech wood spheres pyrolysis under high radiative heat flux wood pyrolysis, respectively.
and a 2-D unsteady numerical model was established to predict the The vapor and pyrolysate leave from the solid surface as soon as
temperature variation and char field. The predicted values can be they are produced, and therefore the solid mass loss rate is taken as
consistent with the experimental results to a certain extent, but it is mass fluxes of volatiles. For the boundary conditions of the front
not accurate enough in predicting the char field. Yuen et al. [10]
established a 3-D model for the pyrolysis of wet wood based on a Table 1
single 1st Arrhenius reaction. This work mainly focused on the Kinetic parameters for pyrolysis of wet wood.
transient pyrolysis of beech wood with different initial moisture
Reaction n A (s1) Ea (J/mol) Ref.
content.
10 4
Overall, these above studies only focused on the effect of Moisture / Vapor 1 5.13 10 8.8 10 [27]
Wood / Char þ Pyrolysate 6.02 5.13 1014 1.63 105 [27]
moisture content on wood pyrolysis. Indeed, moisture in the wood
H. Liu et al. / Energy 201 (2020) 117624 3
X
v v vT
rcp T ¼ k þ u_ i Hi þ q’’net q’’loss (7)
vt vx vx
q’’net ¼ aeff q’’ext εeff s Ts4 T∞
4
hc ðTs T∞ Þ (8)
v v vT 3. Results and discussion
rcp T front ¼ aeff q’’ext þ q’’flame þ k εeff s Ts4 T∞
4
vt vx vx
In our previous study [27], the current numerical model has
hc ðTs T∞ Þ been validated against the experimental data of Shen et al. [24] in
the cone calorimeter in the air atmosphere. The birch as the wood
rYm nm sample with the size of 10 cm 10 cm 1.5 cm and 15.3% moisture
ðrYm Þ0 Am exp
ðrYm Þ0 content was simulated under the fixed radiation heat flux of
Ea;m rYw nw Ea;w 20 kW m2 from the cone calorimeter and 10 kW m2 from the
Hm ðrYw Þ0 Aw exp Hw (10) flame. The results showed that the predicted surface temperature
RT ðrYw Þ0 RT
and total solid conversion ratio agree well with the experimental
data and the numerical model can simulate the wet wood pyrolysis
relatively accurately.
Fig. 1. Mass fluxes of vapor and pyrolysate of 2 mm thickness. Fig. 3. Mass fluxes of vapor and pyrolysate of 12 mm thickness.
4 H. Liu et al. / Energy 201 (2020) 117624
Fig. 5. (a)e(d). Maximum mass fluxes of vapor and pyrolysate under different initial moisture content.
H. Liu et al. / Energy 201 (2020) 117624 5
Fig. 6. (a)e(d). The duration of vapor and pyrolysate under different initial moisture content.
drying process under the action of external heat flux. The free water yield is, which is in accordance with Zeng’s results [5]. The thermal
portion of the total moisture content is consumed during the dry- degradation of wood into tar and then the escape of the tar and
ing process. Then, once the total moisture content of the surface volatiles from wood occur. The increase of initial moisture content
drops to the level close to FSP, the evaporation front moves into the facilitates the wood pyrolysis into more tar [32]. Therefore, the
solid. The moisture exists in the form of vapor behind the evapo- initial moisture content plays a significant role in the wet wood
ration front and the vapor with volatiles flows out of the solid. pyrolysis.
Therefore, the higher initial moisture content generates more va-
pors to mix with volatiles which results in a larger mass fluxes of
3.2. The effect of different moisture gradients
pyrolysate and vapor. Moreover, the duration of the vapor also in-
creases with the increasing initial moisture content. Additionally, it
Usually, the moisture is not homogeneous, but heterogeneous in
is caused by external radiation. When the moisture content is low,
the wood. In this section, the effects of different moisture gradients
the wet wood is quickly heated by external radiation, and then the
on the pyrolysis of wet wood will be discussed. As shown in Fig. 7,
whole time of wet wood pyrolysis decreases.
the birch wet wood with the size of 10 cm 10 cm 1.2 cm was
As mentioned above, the higher the wood moisture content is,
simulated under the fixed radiation heat flux of 40 kW m2 from
the higher the maximum mass fluxes of pyrolysate are. Then, the
the cone calorimeter and 10 kW m2 from the flame. It was
pyrolysates change from the gas phase to the liquid phase. In other
assumed that the moisture was kept in a linear increase or decrease
words, the higher the wood water content is, the higher the liquid
in the current numerical model. The moisture gradient from high to
Fig. 8. (a)e(d). Mass fluxes of vapor and pyrolysate of wet wood with homogenous and heterogeneous moistures.
low was defined as “H-L” while the moisture gradient from low to continuously evaporated out.
high was defined as “L-H”. Four different moisture gradients with It is also found that the mass fluxes of pyrolysate for “H-L” are
the average moisture content 15% were taken into account. much different from those for “L-H”. For the first peak of mass
Fig. 8 (a)-(d) show that the mass fluxes of pyrolysate and vapor fluxes of pyrolysate, when the moisture gradient is “H-L”, a large
of wet wood with homogenous and heterogeneous moisture. It is amount of vapor is first evaporated under the radiation flux
found that the duration of vapor with “H-L” is significantly less than because there is more moisture on the surface. It affects the pro-
that with “L-H” and the maximum mass fluxes with “H-L” are duction of pyrolysate. When the moisture gradient is “L-H”, more
higher than those with “L-H”. This phenomenon becomes more wood is involved in pyrolysis because fewer vapors are released.
prominent when the moisture gradient is wider. The reason is as Therefore, the first peak value for “H-L” is lower than that for “L-H”.
follows: for the wet wood with “H-L”, the radiation flux first heats Note that the second peak value with “H-L” is also lower than that
the high moisture area, and then a large amount of water is evap- for “L-H”. As time goes on, the evaporation front for “H-L” moves
orated out, which results in a large mass fluxes. With the heating from the high moisture area to the low moisture area while the case
proceeding, the evaporation front moves into the low moisture for “L-H” is on the contrary. Thus, there are more vapors with
area, and little water is evaporated out. For the wet wood with “L- volatiles flowing out of the solid for “L-H”, resulting in a higher
H”, the radiation flux first heats the low moisture area and the mass mass fluxes of pyrolysate for “L-H” than those for “H-L”.
fluxes increases slowly. The evaporation moves from the low In order to better study the effects of different moisture gradi-
moisture area to the high moisture area and the moisture is ents on the pyrolysis, the data in Table 2 were plotted in Figs. 9e11.
Table 2
Maximum mass fluxes and duration of vapor and pyrolysate under different moisture gradient.
Moisture gradient Maximum fluxes of vapor (g m2 Duration of vapor Maximum fluxes of pyrolysate of 1st peak (g Maximum fluxes of pyrolysate of 2nd peak (g
(%) s1) (s) m2 s1) m2 s1)
Fig. 9. Maximum fluxes of vapor for “L-H” and “H-L00 Fig. 11. Duration of vapor for “L-H” and “H-L00
Fig. 9 shows that, with the decrease in the moisture gradient, the pyrolysis temperature [33]. Gray et al. [32] also found that more
maximum fluxes of vapor of “L-H” increase while those for “H-L” moisture content results in char yield increase. Therefore, more
decrease. For the maximum fluxes of pyrolysate, as shown in Fig. 10, pyrolysate can be obtained by “L-H” and while “H-L” causes more
two peak values for “L-H” are higher than those for “H-L”. These char. Moreover, Fig. 11 shows that, with the decrease in the mois-
phenomena indicate that the radiation flux first heats the low ture gradient, the duration of vapor for “L-H” decreases while the
moisture area, which leads to more pyrolysate. When the radiation duration of vapor for “H-L” increases. Therefore, the duration of
flux first heats the high moisture area, the moisture vaporization vapor and the maximum fluxes of vapor and pyrolysate for “L-H”
requires more energy to be consumed, which reduces the effective and “H-L” tend to be consistent when the moisture gradient
decreases.
4. Conclusions