Академический Документы
Профессиональный Документы
Культура Документы
www.elsevier.com/locate/ces
Abstract
This paper presents an integrated reaction engineering based mathematical model for clinker formation in cement industry. Separate models
for pre-heater, calciner, rotary kiln and cooler were initially developed and coupled together to build an integrated simulator. Appropriate models
for simulating gassolid contact and heat transfer in pre-heaters were developed. Calciner was modeled by considering simultaneous combustion
of coal particles and calcination of raw meal. Complex heat transfer and reactions (solidsolid, gassolid and homogeneous reactions in gas
phase) in rotary kiln were modeled using three sub-models coupled to each other. Solidsolid reactions in the bed region of the kiln were
modeled using pseudo-homogeneous approximation. Melting of solids in the bed and formation of coating within the kiln were accounted.
Clinker cooler was simulated by developing a two-dimensional model to capture cross-ow heat transfer between air and hot clinkers. The
individual models were coupled with each other via mass and energy communication through common boundaries. The coupled model equations
were solved iteratively. The model predictions agree well with the observations and experience from cement industry. The model was used
to gain better understanding of inuence of operating conditions on energy consumption in cement plant. Several ways for reducing energy
consumption were computationally investigated. The integrated model, the developed software RoCKS (for Rotary Cement Kiln Simulator)
and results presented here will be useful for enhancing our understanding and for enhancing the performance of clinker manufacturing.
2007 Elsevier Ltd. All rights reserved.
Keywords: Cement; Energy consumption; Reaction engineering model
1. Introduction
Cement making processes are extremely energy consuming.
Typically for producing one ton of cement, a well-equipped
plant consumes nearly 3 GJ. For each ton of clinker produced,
an equivalent amount of green house gases are emitted. The
manufacture of cement has been the focus of considerable
attention worldwide because of the high energy usage and high
environmental impact of the process. Considering the recent
impetus on reduction in emission of green house gases and reduction in energy consumption, there is a renewed emphasis
on developing computational models for cement industry and
using this understanding for performance enhancement.
Corresponding author. Tel.: +91 20 2590 2170; fax: +91 20 2590 2621.
Calcineous
Raw meal
2591
Exhaust to
atmosphere
Pre-heater
Assembly
Hot gases
to pre-heater
Coal
Tertiary Air
Calciner
Kiln Exhaust
Pre-heated
Raw meal
Secondary Air
Vent Air
Kiln
Coal
Cooler
Air to cooler
Cooled clinker
2592
height of cyclone = 11 m) indicated that the solid and gas temperatures inside the cyclone were quite uniform (in range of
637644 K) and close to exit temperatures (average temperature of 640 K). Therefore, in this work, each pre-heater cyclone
was treated as completely mixed cell for individual phases and
was, therefore, represented by a pair of temperature (one for
gas phase and the other for solid phase). The heat losses from
cyclone are controlled by the heat ux across the cyclone walls.
Therefore it is essential to predict the heat transfer between
particle laden gas and cyclone walls in the pre-heater cyclone.
The empirical relation proposed by Gupta and Nag (2000) was
used to determine these losses.
2.2. Cyclone calciner
The calcination of raw meal starts in the pre-heater unit itself
once the calcination temperature is reached. In this framework the calcination reaction was assumed to occur in the last
cyclone. The framework is, however, quite general to accommodate relaxation of this assumption. Energy required for the
calcination reaction is given by coal combustion. Shrinking
core reacting model was used for modeling coal combustion.
The char combustion kinetics and gas phase volatile combustion kinetics was used from Hamor et al. (1973) and Li et al.
(2003). The kinetic parameters are given in Table 1. One of the
main issues in modeling cyclone calciner is to estimate the residence time of raw meal in the cyclone. Few empirical correlations have been proposed to predict the average residence time
of particles in cyclones (Kang et al., 1989; Lede et al., 1987).
Pre-heater
Calciner
Strongly
Coupled
Rotary Kiln
Coal/gas phase comb, bed
height variation, clinker formation,
melting/solidification, coating
formation, losses
Clinker Cooler
Gas-solid heat transfer,
temperature gradients in
clinker bed region
Table 1
Kinetic constants for coal combustion reaction
Reactions
K0
E (J/mol)
Reference
C + O2 = CO2
CH4 + 2O2 = CO2 + 2H2 O
6.81 104
1.08 105
2593
wide range of temperatures (2001000 C). Since the temperatures in clinker cooler are in the same range this correlation
was used to model heat transfer coefcient between solids and
gas. The computational models for individual components and
the coupling strategy are discussed in the following section.
3. Computational models and solution methodology
3.1. Cyclone pre-heater model
A schematic of pre-heater unit considered for developing
computational model is shown in Fig. 3a. The present framework of computational models was developed for dry process
of clinker formation since this process is widely used in cement industry. For the dry processes, the moisture content is
generally present in very small amount (typically 0.5%, see
for example Engin and Ari, 2005; Peray, 1984). The energy
requirements for removing the moisture from the feed being
small (less than 0.5% of the total energy consumption), the feed
was considered to be free of moisture in this work. However,
the developed framework is quite general and including evaporation of moisture from the feed is straightforward. The gas
phase and solids in a cyclone was assumed to be completely
back mixed. In Fig. 3a, Ms is the mass of solids entering the
cyclone. Mg is the mass of the air entering the cyclone. Mse is
the mass of solids entrained from a cyclone. Each cyclone was
assumed to be lined with refractory of thickness tr .
Thus, for any ith cyclone in pre-heater assembly the following inlet streams were considered:
1. Solids from the (i 1)th cyclone (Ms,i1 at temperature
Ti1 ).
2. Solids that are entrained by gas from (i + 1)th cyclone
(Mse,i+1 at temperature Ti+1 ).
3. Air from (i + 1)th cyclone (Mg at temperature Ti+1 ).
The outlet streams for this cyclone are:
1. Solids going out of cyclone (Ms,i at temperature Ti ).
2. Solids that are entrained by gas (Mse,i at temperature Ti ).
3. Air going out (Mg at temperature Ti+1 ).
The steady state material balance equation for ith cyclone is
written as
(1)
The hot solids from the kiln are discharged on the grate of
clinker cooler. As the grate moves with uniform speed along
the cooler length, solids lose their heat to cross-ow air. A part
of the air is generally sent to the kiln as secondary air, a part to
calciner as tertiary air and a part is vented to the surroundings
(vent air). The most important key issue in modeling grate coolers is predicting the heat transfer coefcient between hot solids
and cross-ow air. There is no information on modeling of heat
transfer in such cases. In absence of any relevant information
we have used heat transfer correlation in packed bed reactors
to estimate the heat transfer. Nsofor and Adebiyi (2001) have
carried experimental measurements and presented correlation
for forced convection gas particle heat transfer coefcient for
(2)
In the above equations m,p represents the particle capture efciency of the ith cyclone. M represents the mass of the solids
(in kg/s) and subscripts s and se represent solids and entrained
solids, respectively, as explained earlier.
The steady state energy balance for the ith cyclone was written as
Ms,i1 Cp,s Tc,i1 + Mse,i+1 Cp,s Tc,i+1
+ Mg Cp,g Tc,i+1
= Ms,i Cp,s Tc,i + Mse,i Cp,s Tc,i
+ Mg Cp,g Tc,i + hcyc Acyi (Tc,i Tiw,i ).
(3a)
2594
Gas out
Mg,Mse,iTi
Ms,i-1Ti-1
Raw meal
Tiw,i
i
Tow,i
Coal in
Air in
12
Loses
Mg,Mse,i-1Ti-1
1. Refractory
2. Shell
Partially calcined raw meal
Ms,i,Ti
Secondary air
Tertiary air
T
y
Vent air
=0
T
=0
x
Cooling Air, Ta
Fig. 3. (a) Schematic of (a) cement pre-heater, (b) cement calciner, and (c) grate cooler.
In the above Cp,s and Cp,g represents the specic heat of solids
and air, respectively. Subscript g represents the air and Tc,i
represents the temperature of solids and air in the ith cyclone.
hcyc represents the heat transfer coefcient for energy exchange
between particle laden gas and cyclone inner walls. hcyc was
evaluated from the following empirical correlation given by
Gupta and Nag (2000) for heat transfer in cyclones:
hcyc dc
P
= 702.818 + 9.0287 1014 u0 Re + 11.1385
kg
u20
P
+ 4.50398 105
Re + Rc ,
u0
where
Rc = Fpw
4 T4
Tiw
g
Tiw Tg
dc
.
kg
2 L kr [Tiw,i Tr,i ]
,
ln(rr /ri )
(4)
(3b)
The LHS of Eq. (3a) thus represents the total energy entering
the cyclone and RHS represents the energy leaving out of the
cyclone. At steady state the heat given to cyclone walls must
be same as heat conduction in through refractory and cyclone
2595
(11)
The nal term in Eq. (10) represents the loss sensible heat from
coal particle due to loss of mass.
3.2.1.2. Raw meal particle The calcination reaction is given as
(12)
The over all mass balance for the raw meal particle is given by
dmp,L
Mw CO2
,
= rc
Mw CaCO3
d
(13)
where
rc =r ks eE2 /RT L 4 rp2 Mw CaCO3 ,
(14)
(15)
where
(16)
(17)
2596
1
dyv
=
dt
mg
dmCaCO3 ,L
= rc ,
d
(18)
dmCaO,L
rc Mw CaO
=
.
d
Mw CaCO3
(19)
+ [mp,cin yv mp,cout yv ] c Np
rcombg
Vreact Mw vol Zvol
Mw vol
dmg
,
(24)
yvol
dt
dyw
1
mgin ywin mgout ywout
=
mg
dt
rcombg
dmg
Vreact Mw w ZH2 O yw
,
+
Mwvol
dt
(20)
(25)
dmg Cp,g Tg
= mgin Cp,g Tg,in mgout Cp,g Tg
dt
+ Sgcomb + Sccomb + Scalc
dmg
= mgin mgout + [mp,cin mp,cout ] c Np
dt
+ [mp,Lin mp,Lout ]L Np ,
(21)
rcombg
Mwvol
Mwchar
Vreact Mw O2 ZO2 yO2
dmg
dt
+ Cp,c Tcl
dmp,c
d
(1 fc ) Hcomb rcomb
d,
(26b)
mgin yCO2 in mgout yCO2 out
[mp,c yc,cin mp,c yc,cout ] c Np Mw CO2
Mw char
+ [mp,Lin mp,Lout ]L Np
rcombg
Vreact Mw CO2 ZCO2
+
Mw vol
dmg
,
yCO2
dt
(26)
where
Scalc =
(22)
dyCO2
1
=
dt
mg
(23)
In the above equations, Sgcomb, Sccomb are the heat source term
for gas-phase from volatile combustion and char combustion,
respectively. Scalc is the heat sink term from calcination. Hcombg ,
Hcalc are the enthalpies of volatile combustion and calcination.
hc,L is the convective heat transfer coefcient between raw meal
particles and air. The steady state equations across the cyclone
walls were written same as that of pre-heaters explained in
the previous section to obtain temperature of calciner internal
walls, refractory and outer walls.
The calciner model equations were solved using an iterative
method. The model equations for gas phase were solved assuming steady state. For the rst iteration, source terms from
discrete phase were assumed to be zero. The temperature and
2597
(27)
(28)
(29)
j(g ug,y Cp,g Tg )
jy
(31)
2598
It is important to note that the Reynolds numbers for commercial clinker cooler are signicantly higher (Re 1000.2000)
as compared to the experimental conditions of Nsofor and
Adebiyi (2001). However, as discussed earlier, there are no
other systematic experimental studies reported to predict convective heat transfer coefcients in clinker coolers. The empirical correlation (Eq. (31)) was developed for particle sizes close
to those found in industrial clinker coolers and for wide range
of temperature conditions as observed in clinker coolers. Fortunately, the empirical correlation seems to be weekly dependent
on Reynolds number (Reynolds number is to power 0.2). Therefore possible errors associated with Eq. (31) are not expected
to change the simulation results signicantly (predicted Nusselt number is 50.60). Hence Eq. (31) was used to predict
gas solid heat transfer in clinker coolers in the present model.
All the physical properties for determining heat transfer coefcient were calculated at an average temperature of solids and
air as Tf = (Ts + Tg )/2. The system of algebraic linear equations formulated for above model equations was solved using
tri-diagonal matrix algorithm (TDMA).
xCaCO3 ,C =
Mw CO2
,
Mw CaCO3
mCO2 ,C = mCaCO3 ,i P
(32)
(33)
xCaO,C =
mCaCO3 ,i mCaCO3 ,i P
,
Ms,C
mSiO2 ,i
mAl2 O3 ,i
, xAl2 O3 ,C =
,
Ms,C
Ms,C
mFe2 O3 ,i
=
,
Ms,C
xSiO2 ,C =
xFe2 O3 ,C
(34)
(Ms,C ) xCaCO3 ,C
Mw CO2 .
MW CaCO3
(35)
Mw CO2
Mw CaCO3
Mw CO2
,
Mw CaCO3
(36)
where Mg,S is the mass of secondary air entering the kiln, Mg,K
is the air leaving the kiln or entering the calciner, Mc,K is the
amount of coal entering the kiln and yc,K is the mass fraction
of char entering the kiln. The amount of air leaving the preheater assembly was calculated as
Mg,P = Mg,K + Mg,T + mCO2 ,C
+ Mc,C yc,c
Mw CO2
,
Mw CaCO3
(37)
User Input
Dimensions, MOC,
Mass Flow Rate,
Mass Fractions,
Temperature
Call Preprocessor
N
2599
Consistence Checks
&
Generate initial guess
Call Sub-models
Update variables
No
Converged
Yes
Post Processing
(Ms,R Cp,s Ts,R + Mg,T Cp,g Tg,T + Mg,S Cp,g Tg,S ) (Mg,in Cp,g Tg,in )
.
(Ms,K Cp,s )
In the above equation, Mg,in and Tg,in are the mass ow rate
and temperature of air entering the cooler and Ts,R is the temperature of solids exiting the cooler. The temperature of solids
entering the kiln or exiting calciner (Ts,C ) is calculated as
Ts,C =
(39)
were solved by switching off the calcination and clinkerization reactions in the calciner and kiln, respectively. For these
simulations the material and energy balances converged to an
error of 1% giving verication that numerical calculations
are correctly solving the model equations. It was also veried
that the converged solution is not a function of initial guess or
(Ms,C Cp,s Ts,C + Mg,C Cp,g Tg,C + Hcalc Mg,K Cp,g Tg,K Mg,T Cp,g Tg,T Hc,C )
.
(Ms,P Cp,s )
In the above equation, Hc,C is the heat released due to coal combustion in the calciner and Hcalc is the heat required by calcination reaction. This was easily calculated based on percentage
calcination occurring in the calciner. The heat losses in calciner
are negligible as compared to total heat supplied to the calciner
(< 5% of total energy input) and therefore was not considered
in pre-processor calculations. In this way the input conditions
(mass, mass fractions and temperature) for pre-heater, calciner,
kiln and cooler were calculated using pre-processor. The values
calculated by pre-processor were passed as input conditions to
the individual models. The individual models were then solved
iteratively as shown in Fig. 4. The iterations were continued till
(38)
(Ms,K Cp,s Ts,K + Mg,K Cp,g Tg,K + HR,K + HLoss,K Mg,S Cp,g Tg,S Hc,K )
.
(Ms,C Cp,s )
the temperature of solids and gases at exit of individual components were within error of 1%. Suitable under-relaxation
parameters were used. Typically 1020 iterations were required
(40)
2600
Table 3
(a) The physical properties of solids and air, (b) particle size and composition
of coal and (c) particle size and composition of raw meal
Description
Air
Raw meal
Coal
(a)
Thermal conductivity, W/m K
Emmisivity
Heat capacity, J/kg K
Viscoscity, kg/m s
Density, kg/m3
Char caloric value, kcal/kg
Volatile caloric value, kcal/kg
0.116
0.4
1000
1e 05
1.3
0.5
0.9
1000
1500
0.5
0.8
1000
1000
5600
11 900
(b)
Coal particle size, 50 m
Volatile (CH4 )a , 27%
Char, 58%
Ash, 15%
(c)
Raw meal particle size, 50 m
CaCO3 , 80%
CaO, 0%
SiO2 , 14%
Fe2 O3 , 3%
AL2 O3 , 3%
a Mujumdar
Table 2
The dimensions of (a) pre-heater unit, (b) kiln and (c) cooler
S/No.
Description
Units
Values
(a)
1
2
3
4
5
6
7
8
9
10
No. of pre-heaters
Height of cylindrical section
Height of conical section
Diameter of cyclone
Diameter of cone tip
Refractory thickness
Shell thickness
Inlet duct height
Inlet duct width
Diameter of outlet pipe
m
m
m
m
m
m
m
m
m
4
5
3
3
1
0.13
0.03
1
1
1
(b)
1
2
3
4
5
Length
Inner diameter
Coating thickness
Refractory thickness
Shell thickness
m
m
m
m
m
50
3.4
0.136
0.2
0.025
(c)
1
2
Length
Width
m
m
11
1
et al. (2006).
2601
Table 4
Complete energy balance of the system
S/No.
Description
Pre-heater
Calciner
Kiln
Cooler
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
300
50
1214.8
60.8
463.0
2297.9
1650.2
1014.4
98.9
1069.2
50
1114.5
46.7
2.15
350
1650.2
1603.6
23.2
1876.7
1384.5
44.2
1415.3
2297.9
3.22
43.5
1214.8
37.74
1229.9
16.2
0.9
350
1415.3
615.0
9.7
747.1
219.0
1622.4
1603.6
2.24
140.7
1622.4
32.3
300
45
1622.4
416.7
463.0
1415.3
109.4
reasonable when compared with industrial observations. Overall the integrated simulator was able to predict the clinker manufacturing process in cement industry reasonably well.
0.7
CaO -Solid mass fraction
C2S-Solid mass fraction
0.6
Mass fraction
0.5
0.4
0.3
0.2
Calciner
0.1
Cooler
Pre-Heaters
Kiln
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Divisions across the cement clinker process
Fig. 5. Solid and gas mass fractions in pre-heaters, calciner, kiln and cooler
in a cement clinker process.
(41)
Solid temperature
Gas temperature
Gas temperature, K
2000
1500
2000
1500
1000
1000
Solid temperature, K
2602
500
500
Cooler
Calciner
Pre-heaters
Kiln
0
0
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Divisons accross the cement clinker process
Fig. 6. Temperature prole across pre-heaters, calciner, kiln and cooler in a cement clinker process.
1080
635
1060
630
1040
625
1020
620
615
1000
2
4
5
Number of pre-heaters
640
666
1500
661
1400
656
651
1300
646
1200
641
636
1100
1100
Solid temperature entering calciner, K
645
631
626
40
50
60
70
80
% Calcination
90
1000
100
Table 5
Heat of reaction in calciner and kiln
S/No.
Heat of reaction
1
2
Calcination, %
Heat of reaction
in calciner, kJ/kg
clinker
Heat of reaction in
kiln, kJ/kg clinker
50
60
1038.5 1176.6
582.9
433.2
70
1384
219
80
1618.5
90
1802.5
20.2 178.4
630
1500
625
1000
620
615
500
2
5
6
Kiln Rpm
635
1600
630
1200
625
800
620
615
400
1
1.5
2
2.5
Kiln tilt, Degree
1350
645
1300
640
1250
635
1200
630
1150
625
1100
620
0.06
0.08
0.1
0.12
0.14
Grate speed, m/s
0.16
1050
0.18
2000
0.5
650
640
2603
2000
635
2500
640
3.5
Fig. 9. Effect of (a) kiln RPM and (b) kiln tilt on overall energy consumption.
4.2.3. Effect of kiln RPM, kiln tilt and grate speed of clinker
cooler
The effect of kiln rotational speed and kiln tilt on the overall
performance is shown in Fig. 9a and b. For these simulations
the coal ow rate to the kiln was varied to maintain constant
product composition. It can be seen from Fig. 9a as kiln RPM
decreases, the NEC decreases. Changes in kiln RPM changes
the bed height and the residence time of solids in the kiln as
can be seen from Fig. 9a and b (2002.4 s for 3 rpm; 1058.2 s for
5.5 rpm and 703.4 s for 8 rpm). Our simulation results indicate
that it seems to be benecial to operate kilns at lower rpm as
long as adequate mixing of solids is occurring. From Fig. 9b it
can be seen that energy consumptions in kilns operated at lower
tilt is less as compared to kilns at higher tilt. The grate of clinker
cooler is the important parameter that controls the residence
time of solids and subsequently the heat exchange between hot
solids and counter current air in the cooler. We have studied the
inuence of varying grate speed on overall energy consumption.
The results for these simulations are shown in Fig. 10. It can be
seen from Fig. 10 that the NEC increases with increasing grate
speed. The increase in grate speed reduces residence time of
solids in the cooler. This results in less convective heat transfer
between solids and air as clearly indicated by temperature of
secondary air plotted in Fig. 10. Therefore the simulation results
indicate that it is better to operate grates in the cooler at lower
speed. The simulations presented here provide useful trends of
energy consumption as a function of key operating parameters
in cement clinker process. This result also gives us a scope to
understand the importance of design parameters (kiln tilt) on
plant performance and can be very useful to plant engineers.
4.2.4. Effect of solid loading
The predicted results in the form of NEC and corresponding
overall losses for different solids ow rates are shown in Fig. 11.
It can be seen that the NEC per unit weight of product decreases
as solids ow rate increases. This is because the net energy loss
from the entire system decreases as the solid ow rate increases
(see Fig. 11). Thus, it is benecial from the point of view
of energy consumption to operate the units with higher solids
ow rate. Other operational concerns like increase in dusting
and mixing, however, need to be considered while identifying
maximum solids ow rate specically for cement kilns.
4.2.5. Effect of coal composition
The effect of varying coal composition to the kiln on NEC
is shown in Fig. 12. From Fig. 12, it can be seen that the
overall energy consumption does not change signicantly with
changing coal composition (ash content 9%, 15% and 40%).
For these simulations the coal ow rate to the kiln was adjusted so that the same amount of energy is supplied to the
kiln. Therefore the insignicant change in overall energy consumption does not seem to be surprising. However, as the
coal composition changes, the ame characteristics in the kiln
vary. The predicted dimensionless ame length by the simulator for varying coal composition is shown in Fig. 12. The
ame length was calculated by tracking the region in freeboard
where char and volatiles composition in coal go to zero. The
640
330
630
310
620
290
610
270
600
250
44
45
46
47
48
49
Raw meal flow rate, kg/s
50
1120
600
1115
598
1110
596
1105
594
1100
592
1095
590
51
350
650
2604
1090
15
20
25
30
35
40
45
Air flow rate through secondary shell, kg/s
0.56
650
640
0.52
630
0.48
620
0.44
610
600
0.4
30
40
50
60
Char Percentage, %
70
dimensionless ame length was calculated as the ratio of predicted ame length to length of the kiln. It can be seen that
coal with higher ash content tends to have a longer ame as
compared to coal with lower ash content. The ame length is a
complicated function of amount of oxygen, amount of char and
temperature of gas and particle in the freeboard region. Coals
with higher ash content tends to consume oxygen at a slower
rate and therefore result in longer ames. Such simulations can
therefore provide useful information to kiln operators to predict the ame characteristics for wide variety of coal available
in the market.
4.2.6. Effect of secondary shell
Heat losses to the surrounding from the kiln shell by radiation and convection are a signicant source of energy loss in
cement kilns and therefore the overall process. These losses
can be reduced by using a secondary shell. The idea is to cover
the kiln shell with another metallic shell having low surface
hcyc
Hc,C
Hc,K
Hcalc
Hcomb
Hcombg
HLoss,K
HR,K
kg
kr
ks
ksh
ks,c
rk
s
r k
s
L
ma
mAl2 O3 ,i
mCO2 ,C
Notation
a
A0
Acyi
Acyo
Ap
LA
p
Cp,c
Cp,g
Cp,s
dc
dp
E1
E2
fc
fv,0
Fpw
hc
hc,c
hc,L
mCaCO3 ,i
mFe2 O3 ,i
mSiO2 ,i
mg
mgin
mgout
mg,K
mp,c
mp,cin
mp,cout
mpc,0
mp,L
mp,Lin
mp,Lout
ms
Mc,C
Mc,K
Mg
Mg,K
Mg,P
Mg,S
Mg,T
Ms
Mse
Ms,C
Ms,K
2605
2606
Ms,P
MwCaCO3
MwCaO
Mwchar
MwCO2
MwO2
Mwvol
Mww
Nu
cN
p
LN
pO2
pCO2
peq
P
P
Pr
rc
rcomb
rcombg
ri
r0
rp
rr
R
Rc
Re
T0
Tc,i
Tcl
Tf
Tg
Tg,in
Tg,out
Tg,K
Tg,S
Tg,T
Tg,P
Tiw,i
Tow,i
TL
Tr,i
Ts
Ts,C
Ts,R
u0
us,x
us,y
ug,x
ug,y
Vreact
xAl2 O3 ,C
xCaCO3 ,C
xCaO,C
xFe2 O3 ,C
xSiO2 ,C
yc,c
yc,cin
yc,cout
yc,K
yv,c
yO2
yO2 ,in
yO2 ,out
yCO2
yCO2 ,in
yCO2 ,out
yv
yw
Z
Greek letters
c
cy
L
m,p
g
s
Chemical species
C2 S
C3 S
C3A
C4AF
(2CaOSiO2 )
(3CaOSiO2 )
(3CaOAl2 O3 )
(4CaOAl2O3 Fe2 O3 )
Acknowledgments
The authors wish to acknowledge nancial support provided
by CSIR (under the NMITLI scheme) for this study. The authors would also like to acknowledge many helpful discussions
with Professor Anurag Mehra during the course of this work.
One of the authors, K.S.M is grateful to Council of Scientic
and Industrial Research (CSIR), India for providing nancial
support.
References
Engin, T., Ari, V., 2005. Energy auditing and recovery for dry type cement
rotary kiln systemsa case study. Energy Conversion and Management
46 (4), 551562.
Gupta, A.V.S.S.K.S., Nag, P.K., 2000. Prediction of heat transfer coefcient in
the cyclone separator of a CFB. International Journal of Energy Research
24, 10641079.
Hamor, R.J., Smith, I.W., Tyler, R.J., 1973. Kinetics of combustion of
pulverised brown coal char between 630 and 2200 K. Combustion and
Flame 21, 153.
Hindmarsh, A.C., 1983. ODEPACK, a systematized collection of ODE
solvers. In: Stepleman, R.S., et al. (Eds.), Scientic Computing. IMACS
Transactions on Scientic Computation, vol. 1. North-Holland, Amsterdam,
pp. 5564.
Hu, N., Scaroni, A.W., 1996. Calcination of pulverised limestone particles
under furnace injection conditions. Fuel 75, 177186.
Kang, S.K., Kwon, T.W., Kim, S.D., 1989. Hydrodynamic characteristics of
cyclone reactors. Powder Technology 58, 211220.
Kramers, H., Croockewit, P., 1952. The passage of granular solids through
inclined rotarykilns. Chemical Engineering Science 1, 259265.
Lede, J., Li, H.Z., Soulalignac, F., Villermaux, J., 1987. Measurment of solid
particle residence time in cyclone reactor: a comparison of four methods.
Chemical Engineering Process 22, 215222.
Li, J.Q., Wei, F., Jin, Y., 2003. Numerical simulation of pulverized
coal combustion and NO formation. Chemical Engineering Science 58,
51615171.
Locher, G., 2002. Mathematical models for the cement clinker burning
process, part 4: grate cooler. ZKG International 55 (6), 4657.
2607
Lu, J., Huang, L., Hu, Z., Wang, S., 2004. Simulation of gassolid, two-phase
ow, coal combustion and raw meal calcination in a pre-calciner. ZKG
International 57 (2), 5563.
Mastorakos, E., Massias, A., Tsakiroglou, C.D., Goussis, D.A., Burganos,
V.N., 1999. CFD predictions for cement kiln including ame modeling, heat
transfer and clinker chemistry. Applied Mathematical Modeling 23, 5576.
Mujumdar, K.S., Ranade, V.V., 2003. Coupling of CFD models with different
time scalesa case study of rotary cement kiln. In FLUENT Users
Group Meeting held at Pune.
Mujumdar, K.S., Ranade, V.V., 2006. Simulation of rotary cement kilns using
a one-dimensional model. Transactions of IChemE, Part A, Chemical
Engineering Research and Design 84 (A3), 165177.
Mujumdar, K.S., Arora, A., Ranade, V.V., 2006. Modeling of rotary cement
kilns: applications to reduction in energy consumption. Industrial and
Engineering Chemistry Research 45 (7), 23152330.
Nsofor, E.C., Adebiyi, G.A., 2001. Measurements of the gasparticle
convective heat transfer coefcient in a packed bed for high-temperature
energy storage. Experimental Thermal and Fluid Sciences 24 (12), 19.
Peray, K.E., 1984. Rotary Cement Kilns. second ed. Chemical Publishing
Company Ltd., New York.
Spang, H.A., 1972. A dynamic model of a cement kiln. Automatica 8,
309323.
Stanmore, B.R., Gilot, P., 2005. ReviewCalcination and carbonation of
limestone during thermal cycling for CO2 sequestration. Fuel Processing
Technology 86, 17071743.
Warudkar, S., Kumar, G., Mujumdar, K.S., Ranade, V.V., 2005. Modeling of
cement plant calciners. In: CHEMCON 2005, New Delhi.
Watkinson, A.P., Brimacombe, J.K., 1982. Limestone calcination in a rotary
kiln. Metallurgical Transactions B 13B, 369378.