Académique Documents
Professionnel Documents
Culture Documents
Schalk
2010
1
Modelling and Optimisation of a Tunnel Kiln Process
Arnaud Schalk
September, 24
th
2010
Supervised by: Professor Stratos Pistikopoulos
Dr. Kostas Kouramas
Mr. Christos Panos
A thesis submitted to Imperial College London in partial fulfilment of the requirements for the
degree of Master of Science in COURSE TITLE and for the Diploma of Imperial College
Department of Chemical Engineering and Chemical Technology
Imperial College London
London SW7 2AZ, UK
Imperial College London Arnaud
Schalk
2010
2
List of contents
Abstract...........8
Acknowledgments..........9
1. Introduction..........10
2. Literature
review............................................................................................11
2.1 Mathematical Model and Optimization..............................................................11
2.1.1 General Mathematical Model....................................... ...........................11
2.1.2 Temperature fields in tunnel kiln............................................................ 12
2.1.3 Integral Equation......................................................................................12
2.2 Heat transfer coefficient and Experimental studies...........................................13
2.2.1 Heat transfer coefficient...........................................................................13
2.2.2 Experimental Studies...............................................................................13
2.3 Design and Simulation......................................................................................14
2.3.1 Design......................................................................................................14
2.3.2 Simulation................................................................................................15
3 Motivation...............................................................................................................16
4 Process.....................................................................................................................17
4.1 Tunnel Kiln Operation.......................................................................................17
4.2 Characteristic of the firing process....................................................................18
4.3 Process and methodology..................................................................................19
5 Mathematical Model...............................................................................................21
5.1 Assumptions......................................................................................................21
5.2 Model for One Column.....................................................................................22
5.2.1 Variables used in the model.....................................................................22
5.2.2 Properties.................................................................................................26
5.2.3 Properties equations.................................................................................27
5.2.4 Design characteristics..............................................................................28
5.2.5 Mass
Balance.....................................................................................................28
5.2.6 Calculation of the heat transfer coefficient of the gas.............................31
5.2.7 Pressure drop............................................................................................31
5.2.8 Energy Balance........................................................................................32
5.2.9 Heat loss...................................................................................................34
5.3 Parameters assigned for our project..................................................................35
5.4 Index of the model.............................................................................................37
5.5 Initial Conditions...............................................................................................37
Imperial College London Arnaud
Schalk
2010
3
6 Simulation Results for one column.........................................................................39
6.1 Simulation with a fuel Flow rate of 0.02 kg/s...................................................40
6.1.1 Evolution of the gas temperature.............................................................40
6.1.2 Evolution of the Tiles and car temperatures............................................42
6.1.3 Pressure drop............................................................................................44
6.1.4 Heat
loss...........................................................................................................45
6.2 Simulation with a fuel Flow rate of 0.005 kg/s.................................................46
6.2.1 Evolution of the gas temperature.............................................................47
6.2.2 Evolution of the Tiles and car temperatures............................................48
6.2.3 Heat
loss...........................................................................................................48
7 Simulation Results for more than one column........................................................49
Conclusion....................54
Future Work.....................55
References........................56
Appendices....................58
A1. gPROMS Model entity Column.................................................................58
A.2 Model Entity AIR SOURCE......................................................................65
A.3 Model Entity FUEL....................................................................................66
A.4 Model Entity SECONDARY.....................................................................67
A.5 Model Entity MATERIAL.........................................................................68
A.6 Model Entity PROPERTIES......................................................................69
A.7 Model Entity Firing_Column.....................................................................71
A.8 Process Entity Firing Column....................................................................72
A.9 Model Firing Column Topology................................................................73
Imperial College London Arnaud
Schalk
2010
4
List of Figure, Graph and Table
Figure 1: Schematic representation of a Tunnel Kiln.........................................................10
Figure 2: Top schematic view of a tunnel kiln....................................................................17
Figure 3: Picture of two columns of tiles in a Tunnel Kiln...............................................18
Figure 4: Characteristic of the Tunnel kiln..........................................................................18
Figure 5: Example of an experimental temperature distribution......................................19
Figure 6: Theoretical temperature distribution....................................................................20
Figure 7: Side cut of the firing section..................................................................................22
Figure 8: gPROMS one column interface............................................................................39
Figure 9: gPROMS 36 columns interface...........................................................................49
Figure 10: Schematic representation of two connected column......................................50
Graph 1: Inlet temperature of the gas, F
fuel
=0.02 kg/s...................................................40
Graph 2: Outlet temperature of the gas, F
fuel
=0.02 kg/s...................................................41
Graph 3: Tiles temperature, 3D graph, F
fuel
=0.02 kg/s.....................................................42
Graph 4: Tiles temperature, 2D graph, F
fuel
=0.02 kg/s.....................................................42
Graph 5: Car temperature, F
fuel
=0.02 kg/s.........................................................................43
Graph 6: Pressure drop inside the kiln, F
fuel
=0.02 kg/s....................................................44
Graph 7: Heat loss through the walls and roof, F
fuel
=0.02 kg/s.......................................45
Graph 8: Kilns Walls Temperature, F
fuel
=0.02 kg/s........................................................45
Graph 9: Kilns roof Temperature, F
fuel
=0.02 kg/s...........................................................46
Graph 10: Inlet temperature of the gas, F
fuel
=0.005 kg/s..................................................47
Graph 11: Outlet temperature of the gas, F
fuel
=0.005 kg/s...............................................47
Graph 12: Tiles temperature, F
fuel
=0.005 kg/s...................................................................48
Graph 13: Heat loss, F
fuel
=0.005 kg/s.................................................................................48
Graph 14: Tiles temperature, 3D graph, Column 2...........................................................51
Imperial College London Arnaud
Schalk
2010
5
Graph 15: Tiles temperature, 2D graph, Column 2...........................................................52
Graph 16: Car temperature, 2D graph, Column 2.............................................................52
Graph 17: Outlet temperature of the gas, Column 2..........................................................53
Table 1: Model Variables......................................................................................................22
Table 2: Air, Secondary air and Fuel flow rate, temperature, pressure and
composition................................................................................................................................35
Table 3: Inlet temperature car and tiles................................................................................35
Table 4: Design characteristic of the Kiln, Car and Tiles..................................................35
Table 5: Physical properties of the Kiln, Car and Tiles......................................................36
Table 6: Other constants used in the model.........................................................................36
Table 7: Molecular Weight...................................................................................................36
Table 8: First column Inlet.....................................................................................................50
Table 9: Second column Inlet...............................................................................................50
Table 10: Second column Outlet/Third column inlet........................................................51
Imperial College London Arnaud
Schalk
2010
6
Nomenclature
A Area
Cp Heat capacity
D
h
Hydraulic Diameter
f Friction factor
F Flow rate
h Heat convection coefficient
H Enthalpy
i Component i
k Thermal conductivity coefficient
L Length
Height
M MAss
MW Molecular Weight
Nu Nusselt
P Pressure
Per Perimeter
Pr Prandt
Q Heat
r reaction progress
Re Reynolds
Roug Rougness
T Temperature
Thick Thickness
Imperial College London Arnaud
Schalk
2010
7
t Time
u Velocity
V Volume
W Width
x mass fraction
Greek letters
Absorptivity
Emissivity
Density
Stefan-Boltzmann constant
Viscosity
Imperial College London Arnaud
Schalk
2010
8
Abstract:
This work presents the mathematical model developed for the firing section of a tunnel
kiln.
Tunnel kilns are widely used in industry of ceramics and brick making processes. This
is a dynamic process which has not received enough attention in the open literature.
The work presented is mainly based on a model developed in Imperial College
London and on the work made by S. Kaya, E. Mancuhan and K. Kucukada [4] from
Marmara University, Istanbul.
First of all, a model for a unique column of tiles is developed describing all the
physical phenomena happening inside the kiln. This model is then simulated in
gPROMS. Several results based on the simulation are then obtained regarding the
relation between the amount of fuel fed in the tunnel and the inlet temperature of the
gas as well as the temperature distribution inside the tiles.
Secondly, a generalisation of the model for more than one column is attempted and
proved by two separated simulations of one column.
Imperial College London Arnaud
Schalk
2010
9
Acknowledgments
Sincere thanks to Professor Stratos Pistikopoulos and Dr Kostas Kouramas for
their dedicated guidance support throughout the project.
Great thanks to Mr Christos Panos for all his encouragement, great support,
guidance, stimulating suggestions and patience during my project.
Many thanks to all my Imperial College lecturers from whom I gained
invaluable knowledge and without which this thesis would not have been possible.
A particular thanks to Mr Mayank Patel for his thoughtful help during my
project.
Very special thanks to my family for their encouragement and support.
Imperial College London Arnaud
Schalk
2010
10
1. Introduction
Tunnel Kilns are widely used in industry for ceramics and brick making processes. A
definition of tunnel kilns system is given in the U.S. Patent number 4,718,847 of
January, 12
th
, 1988 [1]: [The tunnel Kiln system] is a very long kiln which can be 20
to 100 yards (18m to 92m) in length and through which are continuously moved from
one end to the other a series of carts or cars supporting the materials to be treated. The
kiln is divided into temperature zones varying with distance from the kiln heat source
either by merely utilizing the length of the kiln or by interposing hot gas flow control
systems and movable doors at various locations. In general, a tunnel kiln is composed
of three zones namely (See figure 1): the preheating zone, where bricks are heated to
evaporate the remaining water and to avoid cracking due to the thermal shock, the
firing zone, where the temperature of the bricks is gradually increased to about
1000C, the cooling zone, where heat is recovered.
Figure 1. Schematic representation of a Tunnel Kiln
Many researches and publications have been done about Tunnel Kiln system. We can
identify four main research areas in these studies: The first one is the studies whom
deal with Mathematical Model and Optimization of the process with the objective of
minimizing the costs and the energy. The second one is the calculation of the Heat
transfer coefficient and the analysis of Experimental work. Researches were done
regarding different Design for the tunnel kiln such as small tunnel kiln or rotary tunnel
kiln and also Simulations of tunnel kiln were run via developed software. The last area
is the development of controller in order to run a safe operation, to get the highest
product quality, and to find the most economic operation.
Imperial College London Arnaud
Schalk
2010
11
2. Literature review
2.1 Mathematical Model and Optimization
2.1.1 General Mathematical Model
D.R. Dugwell and D.E. Oakley [2] developed two different models to simulate one
and two dimensional temperature profiles along the tunnel kiln. The model represents
the kiln as a series of plug-flow regions, in which heat transfer to ware occurs,
interspaced by well-stirred adiabatic regions, in which burners and air inleakages are
introduced. The models estimated the gas temperature and composition in addition to
the refractory ware temperature. The two-dimensional form gives good agreement
with measured ware temperature profiles. The simpler one-dimensional form predicts
gas temperature profiles accurately, and estimates representative ware temperatures.
Both models solve the unsteady conduction equation with strongly nonlinear boundary
conditions due to radiation.
In 2005, Ebru Mancuhan and Kurtul Kucukada [3] analysed the operation of a tunnel
kiln producing coal admixed bricks, bricks with a low or high calorific value coal
added as an energy source in the brick body. They considered two different fuels;
pulverized coal (PC) and natural gas (NG), and based their optimization on a particular
plant. The objective was to minimize the cost of the fuels used and the energy lost
through the stack. Solving their 1D model by linear programming, they showed up that
it was advantageous to use admixed coal (AC) for both cases, which are using PC or
NG in the firing zone. The results showed that the using of AC supplies a significant
percentage from 15% to 39% of the energy required for this particular plant. They
showed as well that the optimum energy requirement was between 2040 and 3510
kj/kg brick depending on the properties of the fuels used.
Following this work, Ebru Mancuhan, Kurtul Kucukada and Sinem Kaya [4]
demonstrated how the optimal operating conditions could be predicted by using a
mathematical model representing in the simplest form the phenomena of heat transfer,
combustion of AC and PC, together with gas flow. The work focused on the firing
zone of a tunnel kiln. Using the same 1D-Model and the same objective function than
the previous work, they came up to the conclusion that the heating values of AC and
PC are the key parameters to obtain minimum fuel cost. Using AC with higher heating
values will reduce the required PC while keeping the carbon percentage in the brick
body in the permissible limits and thus the total fuel cost will decrease. They also
determined that the secondary ambient air is blown into the firing zone from 30% to
80% of the dimensionless firing length with a flow rate of 0.175kg/kg while the PC is
fed into the firing zone between the about 18 and 93% of the dimensionless firing
length. The model proposed is specific to coal admixed bricks, and does not consider
the conditions favouring the formation of CO and SO2. The effect of these two
reactions should be included as process constraints in the future.
The same group of research of the Marmara University [5] was interested of the heat
recovery in the cooling zone of a tunnel kiln. As all their previous work, the model
Imperial College London Arnaud
Schalk
2010
12
proposed is still a 1D-model using green bricks. This work was realized to improve the
heat recovery by optimizing the ambient air inlet flow rate at the brick exit side, the
suction and blowing air flow rates along the cooling zone. As the pressure drop along
the cooling zone is a function of the flow rates and temperature of the air, the objective
function to be minimized was defined as the total pressure drop. They determined that
there should be four distinct regions formed of two sets of suction and blowing
regions. This work just gave an estimation of the heat transfer and fluid flow
phenomena and the optimum operating conditions in the cooling zone of a tunnel kiln.
Ebru Mancuhan [6] analysed and optimised the drying of green bricks in a Tunnel
dryer. A drying process is required for the removal of most of the water in the green
brick body to reduce the water content of the brick to about 10% before firing in the
tunnel kiln. Without this preheating step, the water within the clay body turns to
stream at the firing process and damages the bricks severely. The objective function of
this 1D-model is the total cost of the energy required for drying. The cost of the
electricity consumed by the fans for the circulation of the drying air is also included in
the objective function. The optimization of the drying process is realized by linear
programming using Microsoft Excel solver, to find the optimal values of the hot and
outdoor air mass flow rates per kilogram of the brick. The results showed that for
different manipulated variables, it was found that 59 to 62% of the total energy is used
for drying per unit of green bricks.
2.1.2 Temperature fields in tunnel kiln
An approach of a determination of temperature fields in tunnel kiln for brick
production is developed by J. Durakovic and S. Delalic [7]. They claimed their
mathematical model was appropriate for analysis and checking of a stationary
temperature field in brick products and in the furnace. They executed simulations of
temperature distribution in furnace during a brick production process in real conditions
by using a computer program, which has been developed on Delphi 3 programming
language base at Faculty of Metallurgy and Materials Science, University of Zenica.
2.1.3 Integral Equation
In 1985, G. Halasz elaborated a new heat-exchanger model of simulation of heat
treatment in a tunnel kiln. The nonlinear 1D multipoint boundary-value problem has
been handled in the form of an integral equation. Its kernel function characterizing the
mixing and boundary conditions is assumed to consist of a finite number of
elementary kernel functions of its subsystems and the parameters of this kernel
function are directly measurable on a real kiln. A new effective integral equation-
based algorithm solved this equation. Two years later, G. Halasz, J. Toth and K.M.
Hangos [8] applied this model to an existing tunnel kiln in china and proposed a
simple method of determining the energy optimal conditions of this tunnel kiln.
Imperial College London Arnaud
Schalk
2010
13
In 2000, R.H. Essenhigh [9] analysed the tunnel kiln performance to determine the
relation between input energy and useful output energy by applying the integral energy
equation which shown to lead to a firing equation of standard form.
2.2 Heat transfer coefficient and Experimental studies
2.2.1 Heat transfer coefficient:
In 1972, V.G. Abbakumov and G.Sh. Ashkinadze [10] investigated the best
calculations for the coefficient of the heat transfer by convection in tunnel kilns
depending of the structure and the surface of the setting. Based on previous
investigations only, about convective heat exchange from variously orientated
surfaces, and written as an Equations review, they analysed and discussed the
equations to use. They ended up with several recommendations: In calculating the
coefficients of heat transfer by convection in tunnel kilns for the longitudinal surfaces
of the setting with a latticed column structure the following equation is recommended:
Nu=0.08Re
0.7
where
is the Reynolds
criterion; is the heat transmission coefficient, d the equivalent diameter of the
channel; the coefficient of thermal conductivity of the gases; w the velocity of the
gases and v is the kinematic viscosity of the gases. For the longitudinal surfaces of the
setting with solid columns the following expression for the average Nusselt criterion
for the channel is recommended: Nu=0.018Re
0.8
where the coefficient takes into
account the increase in the intensity of the heat exchange with an unstabilized flow in
the setting channels. For transverse surfaces in the setting
kiln
Height of the kiln m 1
L
tile
Length of the tiles column m 1
tiles
Height of the tiles column m 1
M
car
Mass of car kg 1
Imperial College London Arnaud
Schalk
2010
24
M
i
Mass of component i kg 5
M
tiles
Mass of tiles kg 1
M
total
Total Mass kg 1
MW(i) Molecular weight of component i kg/kmol 5
Nu Nusselt Number / 1
P
air_in
Inlet air Pressure Pa 1
P
drop
Drop Pressure Pa 1
Per Perimeter m 1
P
gas_out
Outlet air Pressure Pa 1
Pr Prandt Number / 1
Q
loss
Heat lost J/s 1
Q
radiation
Radiation kJ/s 1
Q
roof
Heat lost through the roof J/s 1
Q
wall
Heat lost through the walls J/s 1
r reaction progress of combustion kmol/s 1
Re Reynolds number / 1
Roug Rougness / 1
T
air_in
Inlet temperature of air K 1
T
amb
Ambient Temperature K 1
T
car_in
Inlet temperature of car K 1
T
car_out
Outlet temperature of car K 1
T
fuel_in
Inlet temperature of fuel K 1
T
gas_in
Inlet temperature of gas K 1
T
gas_out
Outlet temperature of gas K 1
Thick
roof
Thickness of the roof m 1
Thick
wall
Thickness of the walls m 1
Imperial College London Arnaud
Schalk
2010
25
T
roof
Temperature of the roof K 1
T
sec_in
Inlet temperature of secondary air K 1
t
step
Step Time s 1
T
tiles_in
Inlet temperature of the tiles K 1
T
tiles_out
Outlet temperature of the tiles K 1
T
wall
Temperature of the walls K 1
u Velocity of gas m/s 1
V
free
Free volume m
3
1
V
tiles
Volume of the tiles m
3
1
W
step
Width step m 1
W
tiles
Width of the tiles column m 1
x
air_in
Inlet air mass fraction / 5
x
fuel_in
Inlet fuel mass fraction / 5
x
gas_in
Inlet gas mass fraction / 5
x
gas_out
Outlet gas mass fraction / 5
x
sec_in
Inlet secondary air mass fraction / 5
gas
Absorptivity of gas / 1
gas
Emissivity of the gas / 1
kiln
Emissivity of the wall / 1
tiles
Emissivity of the tiles / 1
gas
Density of gas kg/m
3
1
Stefan-Boltzmann constant W/m
2
.K
4
1
gas
Viscosity of gas kg/m.s 1
Total
111
Table 1. Model Variables
Imperial College London Arnaud
Schalk
2010
26
5.2.2 Properties:
The model involves five chemical components: Carbon dioxide, water, oxygen,
nitrogen and methane.
To each of them will be attributed a number used further in the model:
i=1: CO
2
i=2: H
2
O
i=3: O
2
i=4: N
2
i=5: CH
4
Several physical properties of the three different streams are needed in the model.
They depend of the temperature, the pressure and the composition of each stream
considered. Some of them will be calculated by the Physical property packages of
gPROMS[20] and some other will be calculated by correlation from literature.
Enthalpy: Enthalpy is a function of temperature and pressure. Values for the more
common substances have been determined experimentally and are given in the various
handbooks. [21] We will use the package for this property in our model.
H = H(T, P, n)
Density: Enthalpy is a function of temperature, pressure and composition. We will use
the package for this property in our model.
= (T, P, n)
Viscosity: Values for pure substances can usually be found in the literature. It is
function of the temperature, the pressure and the composition when it is for a mixture.
We will use the package for this property in our model.
= (T, P, n)
Heat capacity: It depends of temperature, pressure and the composition of the stream
and will be calculated by the physical properties package.
Cp=Cp(T, P, n)
Thermal conductivity: A correlation will be used. It is the one used by S. Kaya, E.
Mancuhan, K. Kucukada, in Model-based optimization of heat recovery in the
cooling zone of a tunnel kiln [5].
Imperial College London Arnaud
Schalk
2010
27
5.2.3 Properties equations:
Property Equation Number
of
Equation
Equation
Number
Enthalpy of
air
H
air
= H
air
(T
air
, P
air
, F
air
*x
air
(i))
1 1
Enthalpy of
secondary air
H
sec
= H
sec
(T
sec
, P
sec
, F
sec
*x
sec
(i))
1 2
Enthalpy of
fuel
H
fuel
= H
fuel
(T
fuel
, P
uel
, F
fuel
*x
fuel
(i))
1 3
Enthalpy of
gas after
combustion
H
gas
= H
gas
(T
gas_in
, P
gas_in
, F
gas_in
*x
gas_in
(i)) 1 4
Density of the
gas
gas
=
gas
(T
gas_in
, P
gas_in
, F
gas_in
*x
gas_in
(i)) 1 5
Viscosity of
the gas
gas
=
gas
(T
gas_in
, P
gas_in
, F
gas_in
*x
gas_in
(i))
1 6
Heat capacity
of the inlet
gas
Cp
gas_in
=Cp
gas_in
(T
gas_in
, P
gas_in
, F
gas_in
*x
gas_in
(i))
1 7
Heat capacity
of the outlet
gas
Cp
gas_out
= Cp
gas_out
(T
gas_out
, P
gas_out
, F
gas_out
*x
gas_out
(i))
1 8
Thermal
conductivity
1 9
Imperial College London Arnaud
Schalk
2010
28
5.2.4 Design characteristics:
Some design characteristics will be needed for the mathematical model depending of
the characteristics of the kiln and the columns themselves:
5.2.5 Mass Balance:
Gas:
We assume that:
- the mixing of the three inlet streams happens just before getting in the column,
- the reaction of combustion of the fuel happens just before getting in the column,
during the mixing,
- the fuel is fully oxidised (total reaction),
- the inlet gas flow rate and the outlet gas flow rate are constant and even,
- the composition of the inlet gas flow rate is the one after combustion,
- the composition of the outlet gas flow rate is the same than the inlet gas flow rate
because no reaction happens inside the column,
- the pressure of the inlet gas flow rate is the pressure of the inlet air as it is much
bigger than the two others.
Perimeter:
1 10
Area free:
1 11
Area tiles:
1 12
Area car:
1 13
Area walls:
1 14
Area roof:
1 15
Volume free:
1 16
Volume tiles:
1 17
Imperial College London Arnaud
Schalk
2010
29
Before mixing:
Flow rate:
1 18
Composition:
5 19,
20,
21,
22,
23
Reaction:
CH
4
+ 2O
2
CO
2
+ 2H
2
O
From the assumption that the fuel is fully oxidised, we need a condition on the on the
different air flows to keep the excess of oxygen:
We have
The oxygen comes from the air source and the secondary air source:
Therefore, for a viable process we need the fuel flow rate 19.05 times less than the
sum of the two air flows.
Reaction
progress:
1 24
After the mixing and the passage through the column:
Mass
conservation:
1 25
1 26
1 27
1 28
1 29
Imperial College London Arnaud
Schalk
2010
30
The total mass of the gas is the sum of the mass of each component in the gas:
Total mass
1 30
The total mass is also related to the density of the gas:
Total mass
1 31
Composition:
Composition
5 32,
33,
34,
35,
36
Velocity:
Velocity
1 37
Car:
We assume that:
- The mass of the car does not change over the time,
- The car is moving through the tunnel at a constant velocity.
Mass conservation and flow rate:
Mass
1 38
Flow rate
1 39
Constant
flow rate
1 40
Tiles:
We assume that:
- The mass of the tiles does not change over the time,
- No reaction happens inside the tiles material,
- The tiles are moving through the tunnel at a constant velocity.
Imperial College London Arnaud
Schalk
2010
31
Mass conservation and flow rate:
Mass
1 41
Flow rate
1 42
Constant
flow rate
1 43
5.2.6 Calculation of the heat transfer coefficient of the gas:
The heat transfer coefficient of the gas is a function of the temperature of the gas, the
pressure and the velocity of the gas through the column. As the temperature of the gas
is not constant, the heat transfer is not either:
The Nusselt number is a function of the heat transfer coefficient and the hydraulic
diameter:
Nu
1 44
D
h
1 45
It is function of the Reynolds and Prandt Numbers:
5.2.7 Pressure drop:
Due to the friction between the gas and the tiles, a pressure drop can be observed. The
pressure drop needs to be the smallest possible in order to minimize the energy lost
due to the friction.
The pressure drop is a function of the friction coefficient:
P
drop
1 49
1 50
From [3], the pressure drop is described as a differential:
Nu
1 46
Re
1 47
Pr
1 48
Imperial College London Arnaud
Schalk
2010
32
The friction factor is a dimensionless parameter. We assume that the gas is fully
developed turbulent flow; we therefore use the Colebrook equation for the friction
factor.
Friction factor
1 51
5.2.8 Energy Balance:
Combustion:
The heat needed to heat up the tiles comes from the combustion of the methane.
In order to control the firing process we need to know precisely the amount of energy
released by the combustion. It can be done by an enthalpy balance.
The difference between the total inlet enthalpy and the total outlet enthalpy (after
combustion) gives the heat of combustion, or in other words the amount of energy
released by the oxidative reaction. This can be then related to the temperature of the
gas after the reaction, just before getting in the column. The heat released is equal to
the energy needed to increase the temperature of the gas from the temperature at the
reaction took place to the temperature after reaction:
The difference between the inlet temperature (the sum of the three enthalpies of the
three flows) and the outlet enthalpy (just after the reaction took place, during the
mixing) gives the heat of combustion. The standard enthalpy of each component can
easily be found in the literature but as we are operating at high temperature, the
enthalpy is function of the temperature, the composition and the flow rate. All these
parameters are taken in account in the gPROMS physical properties package.
Temperature of
the gas after
combustion
1 52
Gas:
By going through the column, the gas exchanges heat with the tiles to heat them up
and with the car. This heat is transferred by convection and radiation as we are
Imperial College London Arnaud
Schalk
2010
33
operating at very high temperatures. At the same time, some heat is lost by radiation
with the walls and the roof.
The difference of energy is the sum of the heat carried by the gas when it gets in the
column, the heat carried by the gas when it gets out of the column, the energy
transferred by convection to the tiles and the car and the heat lost through the walls
and the roof:
Energy balance
for the gas
1 53
Car:
The difference of heat for the car is only due to the convection between it and the gas:
Energy balance
for the car
1 54
Tiles:
The difference of heat for the tiles is due to the convection between it and the gas and
also by radiation:
Energy balance
for the tiles
1 55
The radiative heat transfered between the gas and a surface (here the tiles) is:
Radiation
1 56
Imperial College London Arnaud
Schalk
2010
34
5.2.9 Heat loss:
According to O.B. Goltsova, V.S. Klekovkin, O.B. Nagovitsin, S.V. Antonychev
[13], the thermal loss has the following form:
Q
loss
=Q
walls
+Q
roof
Where Q
walls
and Q
roof
is the heat lost through the walls and the roof by convection and
radiation.
According to Sinem Kaya, Ebru Mancuan and K. Kucukada [4], the heat loss is
significant in the firing zone. By writing the heat balance equation on the wall and the
roof given by the following equation, the wall and roof temperatures at any location
along the firing zone and the heat loss for the length increment can be calculated:
Heat transfer
by convection between
air and wall/roof
+
Heat transfer to kiln walls
by radiation from gas
=
Heat transfer through the
kiln walls/roof by
conduction
We therefore have the following equations in our model:
Heat losses
1 57
Q
walls
1 58
Q
roof
1 59
Temperature of the walls and temperature of the roof:
T
walls
1 60
T
roof
1 61
Imperial College London Arnaud
Schalk
2010
35
5.3 Parameters assigned for our project:
Several parameters have been assigned for the simulation of our model in gPROMS.
The inlet flow rate, temperature, pressure and composition of the fuel, the secondary
air and the air sources as well as some physical properties that we assume remaining
constant during the process.
The inlet flow rate and temperature of the car and the tiles are also known as well as
the physical properties that we assume remaining constant during the process.
Air, secondary air and fuel inlet:
Air Secondary air Fuel
Flow rate (kg.s
-1
) 0.5 0.2 0.02
Pressure (Pa) 200000 100000 100000
Temperature (k) 800 700 298.15
X
CO2
0 0 0
X
H2O
0 0 0
X
O2
0.21 0.21 0
X
N2
0.79 0.79 0
X
CH4
0 0 1
Table 2. Air, Secondary air and Fuel flow rate, temperature, pressure and composition
Material source:
Tiles Car
Mass 20 20
Temperature 500 500
Table 3. Inlet temperature car and tiles
Dimensions for tiles, car and kiln:
Tiles (Column) Car Kiln Wall Roof
Length (m) 4 Kiln length 4.8 / /
Height (m) 1.9 / 2.3 / /
Width (m) 0.6 W
step
/ / /
Thickness (m) 0.3 0.3
Table 4. Design characteristic of the Kiln, Car and Tiles
Imperial College London Arnaud
Schalk
2010
36
Properties of the Tiles, car and kiln:
Tiles Car Kiln
Heat capacity, Cp (kJ/kg.K) 0.8364 0.7967 /
Heat transfer coefficient
(W/m
2
.K)
/ 1 1
Emissivity coefficient (/) 0.93 / 0.85
Thermal conductivity coefficient
(W/m.K)
/ / 0.75
Table 5. Physical properties of the Kiln, Car and Tiles
Some constant used for the model:
Properties Value
gas
(/) 0.8
air
(/) 1
(Stefan-Boltzmann constant) (W/m
2
.K) 5.67*10
-8
Roug (/) 1
T
step
(s) 180
Table 6. Other constants used in the model
Molecular weight:
Compound Molecular weight (kg/kmol)
CO
2
44
H
2
O 16
O
2
28
N
2
32
CH
4
18
Table 7. Molecular Weight
Imperial College London Arnaud
Schalk
2010
37
5.4 Index of the model:
We have 111 variables:
11 differential variables
100 Algebraic variables
54 Assigned
This leads to 57 unknown variables.
We have 57 equations in our model.
We have 11 algebraic variables so we need 11 initial conditions to get an Index-1
Model.
5.5 Initial Conditions:
We have eleven differential equations in our model; we therefore need eleven initial
conditions:
Equation number (39) and (42), the material mass balance, we need two initial
conditions toward the mass of the car and the mass of the tiles:
M
car
=M
car
(assigned)
M
tiles
=M
tiles
(assigned)
Equation number (26) to (30), the five mass balance for the outlet composition of the
gas, we need four initial conditions toward the initial composition. If we know four of
the five mass fractions, the fifth is also known by the condition:
. We then
need a fifth initial condition. It will be on the inlet flow rate of the gas because the
composition is related to the inlet gas:
x
gas_in
(1)
=
x
gas_out
(1)
x
gas_in
(2)
=
x
gas_out
(2)
x
gas_in
(3)
=
x
gas_out
(3)
x
gas_in
(4)
=
x
gas_out
(4)
F
gas_in
=F
gas_out
(4)
Imperial College London Arnaud
Schalk
2010
38
Equation number (53), the enthalpy balance of the gas, we need an initial condition
toward the inlet temperature of the gas:
T
gas_in
=T
air_in
Equation number (54), (55) and (56) lead us to have three initial conditions on the inlet
temperature of the gas, the car and the tiles:
T
gas_in
= T
gas_out
T
car_in
=T
car_out
T
tiles_in
=T
tiles_out
Imperial College London Arnaud
Schalk
2010
39
6. Simulation Results for one column
After the development of the mathematical model the main part of the project has been
to simulate it on gPROMS. The mathematical model comprises a set of integral,
partial differential and algebraic equations (IPDAEs). gPROMS ModelBuilder 3.3.1 is
a software package for the modeling and simulation of processes combining discrete
and continuous characteristics. gPROMS allows the direct modeling of systems
described by a combination of partial differential equations combined with algebraic
ones [23].
Four this process we separate the column of the three gas sources and of the material.
We consider each source as independent for the simulation. This enables us to
change the inlet parameter easily as shown on figure 8.
The column was the section where we put our mathematical model. With the three
sources separated we can control the inlet Temperature of each streams as well as their
flow rate, composition and inlet pressure (See Appendix A.9)
The purpose of separating all the different section was to facilitate the connection of
more than one column as we will discuss later.
Figure 8. gPROMS one column interface
Imperial College London Arnaud
Schalk
2010
40
6.1 Simulation with a fuel Flow rate of 0.02 kg/s
For our first model we fixed the inlet temperature of the the three gas streams and of
the materials source. The inlet temperature of the air coming from the cooling zone
was fixed at 800K, with a pressure of 1Bar, a flow rate of 0.5 kg/s and the composition
of air (21% O
2
, 79%N
2
). For the secondary air, the inlet temperature was fixed at
700K, the pressure at 1Bar a flow rate of 0.5 kg/s and the composition of air. The fuel
is 100% methane, coming in the tunnel at ambient temperature at 0.02 kg/s and with a
pressure of 1Bar.
The inlet temperature of the car and the tiles is supposed to be equal and of 500K.
(See Appendix A.1, A.2, A.3, A.4, A.5, A.6, A.7 and A.8 for the gPROMS code)
6.1.1 Evolution of the gas temperature
The first thing we can observe after the simulation of the model is that we reach
quickly the required temperature for the inlet gas. The graph 1 shows the evolution of
the temperature over the time. We can see that the combustion takes place
instantaneously and increased the temperature of the gas from 800K to 1126K.
Graph 1. Inlet temperature of the gas, F
fuel
=0.02 kg/s
Imperial College London Arnaud
Schalk
2010
41
On the graph 2 we can observe that the temperature of the outlet gas increases after
passing through the column. It is observed an increase of 4 degrees between the two
temperatures. This is certainly due to the numerical solving and not to a real physical
effect.
It represents an increase of only
.
We can then consider that the inlet and outlet temperature remains constant. It tells us
that the quantity of heat released by the combustion is really important, or that the
amount of heat exchanged and lost is low.
Graph 2. Outlet temperature of the gas, F
fuel
=0.02 kg/s
Imperial College London Arnaud
Schalk
2010
42
6.1.2 Evolution of the Tiles and car temperatures
The simulation shows that the temperature of the tiles increases quickly up to 1043K
in less than 100s. The graph 3 shows the distribution of the temperature in the tiles.
We can see that the temperature distribution inside the column is uniform. There is no
variation of the temperature depending of the height of the column.
Graph 3. Tiles temperature, 3D graph, F
fuel
=0.02 kg/s
The graph 4 shows that it is a really quick process and the desired temperature is
quickly reached.
Graph 4. Tiles temperature, 2D graph, F
fuel
=0.02 kg/s
Imperial College London Arnaud
Schalk
2010
43
In contrast with the tiles, we can observe a different variation of the temperature inside
the car. The temperature increases from 500K to 516K. The radiative exchange
between the gas and the car has not been taken in account in the model. This would
explain the small amount of heat exchanged between the gas and the car (Graph 5). It
is important that we minimize the exchange between the car and the gas. The energy
used to heat up the car is as much as fuel consumed, therefore the car has to be made
with a material with a very low heat transfer coefficient.
Graph 5. Car temperature, F
fuel
=0.02 kg/s
Imperial College London Arnaud
Schalk
2010
44
6.1.3 Pressure drop
The pressure drop was an issue in the process. It needs to be the smallest possible. We
can observe a very small pressure drop inside one column of tiles. We get a pressure
drop of 7.3*10
-3
Pa which is almost null (graph 6). We can say that the pressure drop
after one column is null and the pressure remains the same.
Graph 6. Pressure drop inside the kiln, F
fuel
=0.02 kg/s
Imperial College London Arnaud
Schalk
2010
45
6.1.4 Heat loss
In order to minimize the fuel consumption we need to know how much energy is loss
in the system, through the wall and the roof. For one column we observe a loss of
4030 J/s (Graph 7). This energy is the energy required to heat up the roof and the walls
from 743K to 1060K according to the results of the simulation as shown on graph 8
and 9.
Graph 7. Heat loss through the walls and roof, F
fuel
=0.02 kg/s
Graph 8. Kilns Walls Temperature, F
fuel
=0.02 kg/s
Imperial College London Arnaud
Schalk
2010
46
Graph 9. Kilns roof Temperature, F
fuel
=0.02 kg/s
6.2 Simulation with a fuel Flow rate of 0.005 kg/s
An interesting result is observed while decreasing the amount of fuel fed in the tunnel.
Another simulation has been run with a flow rate of 0.005 kg/s of fuel instead of
0.02kg/s, which is four times less.
We can observe the same quick increase in temperature for the inlet gas flow (Graph
10), up to 1151K which is higher than the temperature got with the previous flow rate.
This means that the process does not need a big amount of fuel to work and that a too
high flow rate would have a negative effect on the process by decreasing the
temperature.
Imperial College London Arnaud
Schalk
2010
47
6.2.1 Evolution of the gas temperature
Graph 10. Inlet temperature of the gas, F
fuel
=0.005 kg/s
It can also be observed that the outlet temperature of the gas is now lower than the
inlet which is now coherent with the fact that the gas exchange heat with the tiles and
the tunnel (Graph 11). We can observe a loss of heat due to the exchange happening in
the column. T
gas_out
=1149K.
Graph 11. Outlet temperature of the gas, F
fuel
=0.005 kg/s
Imperial College London Arnaud
Schalk
2010
48
6.2.2 Evolution of the Tiles and car temperatures
We also observe that the temperature of the tiles is higher than the previous
simulation. With a lower flow rate of fuel, we get a higher inlet gas temperature and a
higher tiles temperature, T
tiles_out
=1061K (Graph 12):
Graph 12. Tiles temperature, F
fuel
=0.005 kg/s
6.2.3 Heat loss
This result is important in order to minimize the fuel consumption.
At the same time, the heat loss remains almost the same (4121 J/s) as shown in the
graph 13:
Graph 13. Heat loss, F
fuel
=0.005 kg/s
Imperial College London Arnaud
Schalk
2010
49
7. Simulation Results for more than one
column
Figure 9. gPROMS 36 columns interface
The general idea with the generalisation of the model was to connect 36 columns
together.
The outlet gas flow rate of the first column was then the inlet gas flow rate of the
second column and so on for the 34 others. It is the same for the material. The outlet
flow rate of material of the 36
th
column was the inlet flow of material for the 35
th
column and so on until the first one.
This is shown on the figure 9.
F
gas_in_2
=F
gas_out_1
and F
tiles_in_35
=F
tiles_out_36
which is the same for the car than the tiles.
To connect the columns together we worked step by step. We could not connect them
all together due to high temperature and large flow rate resulting of the addition of the
column and secondary air/ fuel sources.
Some problems have been encountered while simulating the whole process.
In this project, it has not been possible to run the process with only two connected
columns. A solver issue that we have not been able to fix.
Nevertheless, it has been possible to test if the model could run for more than one
column. For this, the results got after the second simulation (after decreasing the fuel
flow rate) have been implemented for a new simulation as the input of the new
simulation (figure 10):
Imperial College London Arnaud
Schalk
2010
50
Column
1
Fuel
F
fuel
=0.005 kg/s
T
fuel
=298.15K
Secondary air
F
sec
=0.05 kg/s
T
sec
=700K
Air
F
airc
=0.05 kg/s
T
air
=800K
F
gas_out_1
=F
gas_in_2
T
gas_out_1
=T
gas_in_2
x(i)
gas_out_1
=x(i)
gas_in_2
Column
2
F
gas_out_2
=F
gas_in_3
T
gas_out_2
=T
gas_in_3
x(i)
gas_out_2
=x(i)
gas_in_3
Fuel
F
fuel
=0.005 kg/s
T
fuel
=298.15K
Secondary air
F
sec
=0.05 kg/s
T
sec
=700K
Figure 10. Schematic representation of two connected column
First simulation (Table 8):
F
gas_in_1
=F
air_1
F
fuel_1
F
sec_1
Flowrate (kg/s) 0.5 0.005 0.5
Temperature (K) 800 298.15 700
Pressure (Pa) 100000 100000 100000
x(1) 0 0 0
x(2) 0 0 0
x(3) 0.21 0 0.21
x(4) 0.79 0 0.79
x(5) 0 1 0
Table 8. First column Inlet
We got, after the first simulation the following results (Table 9):
F
gas_in_2
=F
gas_out_1
F
fuel_2
F
sec_2
Flowrate (kg/s) 1.005 0.005 0.5
Temperature (K) 1150 298.15 700
Pressure (Pa) 99999.99 100000 100000
x(1) 0.014 0 0
x(2) 0.01 0 0
x(3) 0.19 0 0.21
x(4) 0.786 0 0.79
x(5) 0 1 0
Table 9. Second column Inlet
Imperial College London Arnaud
Schalk
2010
51
We then obtain the following results after the second simulation for the second
column, with a random temperature for the tiles and the car of: T
tiles
=1000K and
T
car
=516 (Table 10):
F
gas_in_3
=F
gas_out_2
F
fuel_3
F
sec_3
Flowrate (kg/s) 1.51 0.005 0.5
Temperature (K) 1334 298.15 700
Pressure (Pa) 99999.98 100000 100000
x(1) 0.018 0 0
x(2) 0.014 0 0
x(3) 0.183 0 0.21
x(4) 0.785 0 0.79
x(5) 0 1 0
T
tiles
(K) 1255
T
car
(K) 537
Table 10. Second column Outlet/Third column inlet
This simulation shows that with constant values for the inlet flow, we can run the
simulation. As it is a dynamic simulation, the parameters of the new inlet flows are
variables and change during the simulation, and it seems to be the problem of the fail
of the simulation.
From the following graph (Graphs 14, 15, 16, 17), the results are coherent with the
simple simulation in point (5.2):
Graph 14. Tiles temperature, 3D graph, Column 2
Imperial College London Arnaud
Schalk
2010
52
Graph 15. Tiles temperature, 2D graph, Column 2
Graph 16. Car temperature, 2D graph, Column 2
Imperial College London Arnaud
Schalk
2010
53
Graph 17. Outlet temperature of the gas, Column 2
Imperial College London Arnaud
Schalk
2010
54
Conclusion
This work focuses mainly on the development of a stable mathematical model for the
firing section of the tunnel kiln. Based on previous work and assumptions, a dynamic
model was first of all developed for one column of tiles going through the tunnel.
The model describes all the physical phenomena inside de kiln such as the combustion
of the fuel, the heat transfer by convection and radiation between the tiles and the gas,
the heat transfer loss through the walls and roof by convection and radiation, the
pressure drop due to the friction of the gas with the tiles and the mass balance for the
gas and the materials.
By simulating this model with gPROMS, it was shown that the model was stable and
coherent. By several simulations, it has been shown that the amount of fuel fed in the
tunnel was a key point for a proper cooking of the tiles. It was shown that a too high
amount of fuel could lead to an effect which is against the effect desired: decreasing
the temperature of the inlet gas.
A simulation for the whole process was attempted without success. In order to prove
that the model could work for more than one column, two separated simulations were
run. It proved that the model could run and give coherent results. An increased of
temperature was observed for the gas, the tiles and the car.
Regarding the optimisation of the process, it was planned to run an optimisation in
order to minimize the fuel consumption. As the simulation of the whole process has
not worked, the optimisation has not been done. It would be a future work.
Imperial College London Arnaud
Schalk
2010
55
Future Work
According to what is mentioned above, the project concentrates on the modelling of a
stable dynamic model for the firing zone of a tunnel kiln, and the simulation of this
model in gPROMS. Several results have been obtained for one column going through
the tunnel.
However the simulation for more than one column has not been possible in this
project. The model seems to work but it needs more attention to make it work for two
columns and then the whole process with its 36 columns.
Additionally, after the simulation of the whole process, the optimization of the system
in order to minimize the fuel consumption should be run.
Imperial College London Arnaud
Schalk
2010
56
References:
[1] United States Patent, Kiln System, Patent number 4,718,847; Jan. 12, 1988.
[2] D.R. Dugwell, D.E. Oakley, A model of heat transfer in tunnel kilns used for firing
refractories, Int. J. Heat Mass Transfer 1988, 31:2381-2390
[3] E. Mancuhan, K. Kucukada, Optimization of fuel and air use in a tunnel kiln to
produce coal admixed bricks, Appl Therm Eng 2006; 26:1556-1563
[4] S. Kaya, E. Mancuhan, K. Kucukada, Modelling and optimization of the firing
zone of a tunnel kiln to predict the optimal feed locations and mass fluxes of the fuel
and secondary air, App Energy 2009; 86:325-332
[5] S. Kaya, E. Mancuhan, K. Kucukada, Model-based optimization of heat recovery
in the cooling zone of a tunnel kiln, App Therm Eng 2008; 28:633-641
[6] E. Mancuhan, Analysis and optimization of drying of green bricks in a tunnel
dryer, Drying Technology; 27:5:707-713
[7] J. Dukakovic, S. Delalic, Temperature field analysis of tunnel kiln for brick
production, Materials and Geoenvironment 2006, Vol 53, 3:403-408
[8] G. Halasz, J. Toth and K.M. Hangos, Energy-optimal operation conditions of a
tunnel kiln, Comput. Chem. Engng 1988, 12:183-187
[9] R.H. Essenhigh, Studies in Furnace Analysis: Prediction of Tunnel Kiln
Performance by Application of the Integral Energy Equation, Energy and fuels 2001,
15 :552-558
[10] V.G. Abbakumov, G.Sh. Ashkinadze, Convective Heat Exchange in Tunnel Kiln,
Refractories and Industrial Ceramics 1972; 13:3:20-27
[11] G.A. kovelman, A.A. Barenboim, Heat exchange during heating of fine ceramics
in tunnel kilns, State Institute of Ceramics Industry 1974, 10:20-21
[12] S.A. Karaush, Yu. I. Chizhik, E.G. Bober, Optimization of ceramic setting as a
function of their heat absorption from the radiating walls of the furnace, Steklo i
Keramika 1997, 6:25-27
Imperial College London Arnaud
Schalk
2010
57
[13] O.B. Goltsova, V.S. Klekovkin, O.B. Nagovitsin, S.V. Antonychev, Heat losses
in a tunnel kiln for brick firing, Steklo i Keramika 2006, 4:24-25
[14] A. F. Utenkov, E. A. Sinitsyn, V. G. Abbakumov, M. S. Glazman, V. F.
Konstantinov, Yu.F. Sachkov, L. I. Krigman, A. A. Ionin, and B. A. Unaspekov ; A
system for firing a Small tunnel kiln operating on natural gas , Ogneupory 1983, 3 :33-
39
[15] V. G. Abbakumov, G. A. Taraka~chikov, S. I. Vel'sin, Yu. G. Golod, E. I.
Telkman, E. A. Drozdov, A. A. Kulikov, A. G. Belogrudov, I. V. Zimnukhov, N. A.
Domrachev, and A. S. Potapov ; A circular tunnel kiln for firing of refractories,
Ogneupory 1985, 2 :40-44
[16] N.A. Tyutin, B.I. Kitaev, V.G. Avdeeva, Investigation of the aerodynamics of a
tunnel circular kiln, Orgneupory 1982, 6:20-27
[17] J.F.M. Vellhuis, J. Denissen, Simulation model for industrial dryers: reduction of
drying times of ceramics and saving energy, Drying technology, 15:6, 1941-1949
[18] O.B. Goltsova, V.S. Klekovkin, O.B. Nagovitsyn, N.L. Dmitriev, Cause-and-
effect relations with respect to defects in brick firing in tunnel kilns, Steklo I Keramika
2005, 3:26-28
[19] P. Michael, S. Manesis, Modelling and control of industrial tunnel-type furnaces
for brick and tile production. Proceeding of the5th international conference on
technology and automation. Greece: Thessaloniki; 2005. p. 21621.
[20] gPROMS Physical Properties Guide
[21] R.K. Sinnott, Chemical Engineering Design Volume 6, Coulson and Richardsons
Chemical Engineering Series, Fourth Edition
[22] Chemical Engineering Department, MSc Handbook, Imperial College London
2009
[23] PSE, http://www.psenterprise.com/
Imperial College London Arnaud
Schalk
2010
58
APPENDICES
A1. gPROMS Model entity Column
PARAMETER
# Number of components
NoComp AS INTEGER
phys_prop AS FOREIGN_OBJECT "PhysProp"
Tileheight AS REAL
DISTRIBUTION_DOMAIN
Y AS [ 0 : Tileheight ] # normalised domain
UNIT
properties as PropertiesTileCar
Material_Source001 AS Material_Source
PORT
Air_Inlet AS Air_Flowrate
Air_Outlet AS Air_Flowrate
Fuel_Inlet AS Fuel_Flowrate
Secondary_Inlet AS Sec_Flowrate
Material_Inlet AS Material_Flowrate
Material_Outlet AS Material_Flowrate
VARIABLE
#Flowrate
Fair_in as MassFlowrate
Ffuel_in as MassFlowrate
Fsec_in as MassFlowrate
Fgas_in, Fgas_out as MassFlowrate
Mtotal as Mass
M as array(NoComp) of Mass
FcarIn, FcarOut as MassFlowrate
Mcar as Mass
FtilesIn, FtilesOut as MassFlowrate
Mtiles as Mass
xair_in as array(NoComp) of NoType
xfuel_in as array(NoComp) of NoType
xsec_in as array(NoComp) of NoType
Imperial College London Arnaud
Schalk
2010
59
xgas_in, xgas_out as array(NoComp) of NoType
# Combustion
r as MolarFlowrate
#Velocity
velocity as Velocity
# No-type constants
# Time constants
StepTime as notype
# Characteristic Numbers
Re as notype
Pr as notype
Nu as notype
friction as friction
# General Variables
Combined_variable as notype
Dh as Length
Per as area
Afree as area
Atiles as area
Acar as area
Vfree as volume
Vtiles as volume
Awall as area
Aroof as area
#Pressure
Pair_in as Pressure
Pgas_out as Pressure
P_drop as Pressure
Psec_in, Pfuel_in as Pressure
#Heat
QLosses as Heat
Qwall as Heat
Qroof as Heat
Imperial College London Arnaud
Schalk
2010
60
#S as notype
#Temperature
Tair_In as Temperature
Tfuel_In as Temperature
Tsec_In as Temperature
Tgas_In, Tgas_out as Temperature
TcarIn as Temperature
TcarOut as Temperature
TtilesIn as DISTRIBUTION(Y) OF Temperature
TtilesOut as DISTRIBUTION(Y) OF Temperature
Twall as Temperature
Troof as Temperature
#Heat and Radiation coeff
h_convection as convection
k_conductivity as conduction
#Radtiles as heat
# properties
MolecularWeight as array(NoComp) of molecular_weight
density_air as density
viscosity_air as viscosity
cp_gas_out as HeatCapacity
cp_gas_In as HeatCapacity
#Enthalpy
H_in1,H_in2,H_in3, H_out as Heat
random as Heat
set
Phys_prop :=
"IPPFO::mass:<CARBON_DIOXIDE,WATER,OXYGEN,NITROGEN,METHANE>
" ;
NoComp := Phys_prop.NumberOfComponents ;
Air_Inlet.no_components := NoComp;
Air_Outlet.no_components := NoComp;
Fuel_inlet.no_components := NoComp;
Secondary_inlet.no_components := NoComp;
Tileheight := 1.9;
Y := [ BFDM, 1, 50 ];#[ OCFEM, 2, 5 ] ;
Imperial College London Arnaud
Schalk
2010
61
BOUNDARY
EQUATION
#Connection Variables
# AIR INLET
Air_Inlet.info_air_pressure=Pair_in;
Air_Inlet.info_air_mass=Fair_in;
Air_Inlet.info_air_temperature=Tair_In;
Air_Inlet.info_air_x()=xair_in();
#Fuel Inlet
Fuel_Inlet.info_fuel_pressure=Pfuel_in;
Fuel_Inlet.info_fuel_mass=Ffuel_in;
Fuel_Inlet.info_fuel_temperature=Tfuel_In;
Fuel_Inlet.info_fuel_x()=xfuel_in();
#Secondary air Inlet
Secondary_Inlet.info_sec_pressure=Psec_in;
Secondary_Inlet.info_sec_mass=Fsec_in;
Secondary_Inlet.info_sec_temperature=Tsec_In;
Secondary_Inlet.info_sec_x()=xsec_in();
#Air OUTLET
Air_Outlet.info_air_mass=Fgas_out;
Air_Outlet.info_air_pressure=Pgas_out;
Air_Outlet.info_air_temperature=Tgas_Out;
Air_Outlet.info_air_x()=xgas_out();
#Material INLET
Material_Inlet.info_car_temperature=TcarIn;
FOR z := 0| TO Tileheight| DO
Material_Inlet.info_tile_temperature(z)=TtilesIn(z);
end
#Material OuTLET
Material_Outlet.info_car_temperature=TcarOut;
FOR z := 0| TO Tileheight| DO
Material_Outlet.info_tile_temperature(z)=TtilesOut(z);
end
Imperial College London Arnaud
Schalk
2010
62
#Properties
1000*H_in1*Fair_in=phys_prop.Enthalpy(Tair_In,Pair_in,Air_Inlet.info_air_x*Fair_i
n);
1000*H_in2*Fsec_in=phys_prop.Enthalpy(Tsec_In,Psec_in,Secondary_Inlet.info_sec
_x*Fsec_in);
1000*H_in3*Fsec_in=phys_prop.Enthalpy(Tfuel_In,Pfuel_in,Fuel_Inlet.info_fuel_x*
Ffuel_in);
1000*H_out*Ffuel_in=phys_prop.Enthalpy(Tgas_In,Pair_in,xgas_out*Ffuel_in);
random=-H_out+H_in1+H_in2+H_in3;
density_air=phys_prop.VapourDensity(Tgas_Out, Pgas_out, xgas_out*Mtotal);
viscosity_air=phys_prop.VapourViscosity(Tgas_Out, Pgas_out, xgas_out*Mtotal);
(1000*Mtotal)*cp_gas_In=phys_prop.VapourHeatCapacity(Tgas_in, Pair_in,
xgas_in*Mtotal);
(1000*Mtotal)*cp_gas_Out=phys_prop.VapourHeatCapacity(Tgas_Out, Pgas_out,
xgas_out*Mtotal);
1000*k_conductivity=4/10^8*(Tgas_out)^3-1/10^4*(Tgas_out)^2 + 0.14*(Tgas_out)-
6.41 ;
MolecularWeight=phys_prop.MolecularWeight;
#Design characteristic
Per=2*(properties.TileLength+properties.Tileheight);
Afree=properties.kilnlength*properties.kilnheight-Atiles;
Atiles=properties.TileLength*properties.Tileheight;
Acar=properties.kilnLength*properties.WidthStep;
Vfree=properties.kilnlength*properties.kilnheight*properties.WidthStep-Vtiles;
Vtiles=Atiles*properties.TileWidth;
Awall=2*(properties.WidthStep*properties.kilnheight);
Aroof=properties.kilnlength*properties.WidthStep;
#Mass balance gas
Fgas_in=Fair_in+Ffuel_in+Fsec_in;
r=(Fgas_in*xgas_in(5))/properties.MolCH4;
for i:=1 to NoComp do
Fgas_in*xgas_in(i)= Fair_in*xair_in(i)+Ffuel_in*xfuel_in(i)+Fsec_in*xsec_in(i);
end
$M(1)=Fgas_in*xgas_in(1)-Fgas_out*xgas_out(1)+r*properties.MolCO2;
$M(2)=Fgas_in*xgas_in(2)-Fgas_out*xgas_out(2)+2*r*properties.MolH2O;
$M(3)=Fgas_in*xgas_in(3)-Fgas_out*xgas_out(3)-2*r*properties.MolO2;
Imperial College London Arnaud
Schalk
2010
63
$M(4)=Fgas_in*xgas_in(4)-Fgas_out*xgas_out(4)+0*r*properties.MolN2;
$M(5)=Fgas_in*xgas_in(5)-Fgas_out*xgas_out(5)-r*properties.MolCH4;
Mtotal=SIGMA(M());
Mtotal=Vfree*density_air;
For i:=1 to NoComp do
Mtotal*xgas_out(i)=M(i);
END
Fgas_out=Afree*density_air*velocity;
#Heat transfer coeff
Per*Dh=4*Atiles;
viscosity_air*Re=velocity*Dh*density_air;
k_conductivity*Pr=1000*cp_gas_Out*viscosity_air;
k_conductivity*Nu=h_convection*Dh;
Nu=0.021*Re^(0.8)*Pr^(1/3);
#Pressure drop:
1=-2*log10(2.51/((Re*friction^(1/2)))+1/(3.7*Dh))*friction^(1/2);
2*Dh*P_drop=friction*(density_air*velocity^2*properties.WidthStep);
P_drop=Pair_in-Pgas_out;
#Mass Balance for tiles:
$Mtiles=FtilesIn-FtilesOut;
FtilesIn=Mtiles/properties.StepTime;
FtilesIn=Ftilesout;
#Mass Balance for car
$Mcar=FcarIn-FcarOut;
properties.StepTime*FcarIn=Mcar;
FcarIn=FcarOut;
#Energy Balances:
Fgas_in*cp_gas_In*$Tgas_In=H_in1+H_in2+H_in3-H_out;
#Gas
Mtotal*$(combined_variable)=Fgas_in*cp_gas_In*Tgas_In -
Fgas_Out*cp_gas_Out*Tgas_Out - properties.hcar*Acar*(Tgas_Out-TcarOut)/1000
- h_convection*Atiles*integral(z:=0:TileHeight;(Tgas_Out-
TtilesOut(z)))/Tileheight/1000 -QLosses/1000 ;
Imperial College London Arnaud
Schalk
2010
64
combined_variable=cp_gas_Out*Tgas_Out;
#Car
Mcar*properties.cp_car*$TcarOut=properties.cp_car*(FcarIn*TcarIn -
FcarOut*TcarOut)
+ properties.hcar*Acar*(Tgas_Out-TcarOut)/1000;
#Tiles
for z := 0 TO Tileheight DO
Mtiles*properties.cp_tiles*$TtilesOut(z)=
properties.cp_tiles*(FtilesIn*TtilesIn(z) - FtilesOut*TtilesOut(z))
+ h_convection*Atiles*(Tgas_Out-TtilesOut(z))/1000 +
properties.SB_constant*Atiles*properties.Emissivity_tiles*(properties.Emissivity_air*
Tgas_Out^4 - properties.Absorptivity_air*TtilesOut(z)^4)/1000;
end
#Heat losses
Qlosses=Qwall + Qroof;
Qwall= Awall*(properties.hwall*(Tgas_Out-Twall))+
properties.SB_constant*properties.Emissivity_wall*(properties.Emissivity_air*Tgas_
Out^4 - properties.Absorptivity_air*Twall^4);
Qroof=Aroof*(properties.hwall*(Tgas_Out-
Troof))+properties.SB_constant*properties.Emissivity_wall*(properties.Emissivity_ai
r*Tgas_Out^4 - properties.Absorptivity_air*Troof^4);
#Twall
properties.kw*Awall*((Twall-properties.Temperature_amb)/properties.Thickwall)=
properties.hwall*Awall*(Tgas_Out-Twall) +
properties.SB_constant*Awall*properties.Emissivity_wall*(properties.Emissivity_air
*Tgas_Out^4 - properties.Absorptivity_air*Twall^4);
#Troof
properties.kw*Aroof*((Troof-properties.Temperature_amb)/properties.Thickroof)=
properties.hwall*Aroof*(Tgas_Out-Troof)+
properties.SB_constant*Aroof*properties.Emissivity_wall*(properties.Emissivity_air
*Tgas_Out^4 - properties.Absorptivity_air*Troof^4);
StepTime=properties.StepTime;
ASSIGN
Imperial College London Arnaud
Schalk
2010
65
INITIAL
Tgas_In=Tair_In;
Mcar=20;
Mtiles=20;
xgas_in(1)=xgas_out(1);
xgas_in(2)=xgas_out(2);
xgas_in(3)=xgas_out(3);
xgas_in(4)=xgas_out(4);
Fgas_in=Fgas_Out;
TcarOut=TcarIn;
Tgas_Out=Tgas_In;
FOR z := 0| TO TileHeight| DO
TtilesOut(z)=TtilesIn(z);
END
A.2 Model Entity AIR SOURCE
PARAMETER
NoComp as INTEGER
PORT
Air as Air_Flowrate
PORTSET
Variable
PressureInlet as pressure
AirMassflowrate as MassFlowrate
AirMassFraction as array(NoComp) of notype
AirTemprature as Temperature
set
NoComp := 5 ;
Air.no_components:=NoComp;
equation
air.info_air_mass = AirMassflowrate;
air.info_air_x = AirMassFraction;
air.info_air_temperature = AirTemprature;
air.info_air_pressure = PressureInlet;
Imperial College London Arnaud
Schalk
2010
66
ASSIGN
#PressureInlet:=200000;
#AirTemprature:=800+273.15;
#AirMassflowrate:=3;
#AirMassFraction(1):=0.2;
#AirMassFraction(2):=0.2;
#AirMassFraction(3):=0.2;
#AirMassFraction(4):=0.4;
#AirMassFraction(5):=0.0;
A.3 Model Entity FUEL
PARAMETER
NoComp as INTEGER
PORT
Fuel as Fuel_Flowrate
PORTSET
Variable
PfuelInlet as pressure
FuelMassflowrate as MassFlowrate
FuelMassFraction as array(NoComp) of notype
FuelTemperature as Temperature
set
NoComp := 5 ;
Fuel.no_components:=NoComp;
EQUATION
fuel.info_fuel_pressure = PfuelInlet;
fuel.info_fuel_x = FuelMassFraction;
fuel.info_fuel_mass = FuelMassflowrate;
fuel.info_fuel_temperature = FuelTemperature;
ASSIGN
#PfuelInlet := 100000;
#FuelTemperature := 25 + 273.15;
#FuelMassflowrate := 0.5;
#FuelMassFraction(1):=0;
#FuelMassFraction(2):=0;
Imperial College London Arnaud
Schalk
2010
67
#FuelMassFraction(3):=0;
#FuelMassFraction(4):=0;
#FuelMassFraction(5):=1;
A.4 Model Entity SECONDARY
PARAMETER
NoComp as INTEGER
PORT
Secondary as Sec_Flowrate
PORTSET
Variable
PsecInlet as pressure
SecMassflowrate as MassFlowrate
SecMassFraction as array(NoComp) of notype
SecTemperature as Temperature
set
NoComp := 5 ;
Secondary.no_components:=NoComp;
EQUATION
Secondary.info_sec_pressure = PsecInlet;
Secondary.info_sec_x = SecMassFraction;
Secondary.info_sec_mass = SecMassflowrate;
Secondary.info_sec_temperature = SecTemperature;
ASSIGN
#PsecInlet := 100000;
# SecTemperature := 300 + 273.15;
#SecMassflowrate := 0.5;
#SecMassFraction(1):=0;
#SecMassFraction(2):=0;
#SecMassFraction(3):=0.21;
#SecMassFraction(4):=0.79;
#SecMassFraction(5):=0;
Imperial College London Arnaud
Schalk
2010
68
A.5 Model Entity MATERIAL
PARAMETER
Tileheight as REAL
DISTRIBUTION_DOMAIN
Y AS [ 0 : Tileheight ] # normalised domain
PORT
Material as Material_Flowrate
PORTSET
Variable
Temperature_car as Temperature
Temperature_tiles as DISTRIBUTION (Y) OF Temperature
SET
Tileheight:=1.9;
Y := [ BFDM, 1, 50 ];
equation
Material.info_car_temperature=Temperature_car;
Material.info_tile_temperature()=Temperature_tiles();
ASSIGN
#FOR z := 0 TO Tileheight DO
# Temperature_tiles(z):=700;
#end
# Temperature_car:=600;
Imperial College London Arnaud
Schalk
2010
69
A.6 Model Entity PROPERTIES
PARAMETER
no_components as integer default 5
#co2,h20,02,n2,co
variable
cp_car as HeatCapacity #default 0.7967
cp_tiles as HeatCapacity
hcar as HeatCoef
StepTime as notype
Temperature_amb as Temperature
TileLength as Length
Tileheight as Length
TileWidth as Length
WidthStep as Length
kilnlength as Length
kilnheight as Length
RoofWidth as Length
WallWidth as Length
Emissivity_tiles as NoType
Emissivity_wall as NoType
SB_constant as NoType
Emissivity_air as NoType
Absorptivity_air as Notype
hwall as Heatcoef
Thickwall as Length
Thickroof as Length
cp_air as HeatCapacity
cp_fuel as HeatCapacity
cp_sec as HeatCapacity
#Gas molecular weight
MolCO2 as NoType
MolCH4 as NoType
MolN2 as NoType
MolO2 as NoType
MolH2O as NoType
assign
#general values
StepTime:=180;
Temperature_amb:=273+20;
#air properties
Emissivity_air:=0.8;
Imperial College London Arnaud
Schalk
2010
70
Absorptivity_air:=1;
#tiles properties
cp_tiles := 836.4/1000; #kJ/kg K
TileLength := 4;
Tileheight := 1.9;
TileWidth := 0.6;
WidthStep := 18/6/6;
Emissivity_tiles := 0.93;
SB_constant := 5.67E-8;
#Kiln prop
RoofWidth := 3;
WallWidth := 3;
Thickwall := 0.3;
Thickroof := 0.3;
kilnlength := 4.8;
kilnheight := 2.3;
Emissivity_wall := 0.85;
hwall := 1;
kw := 0.75;
# Car
Ac:=8.496;
kc:=0.7;
hcar:=1;
cp_car:=0.7967;
MolCO2:=44;
MolCH4:=16;
MolN2:=28;
MolO2:=32;
MolH2O:=18;
Imperial College London Arnaud
Schalk
2010
71
A.7 Model Entity Firing_Column
UNIT
Air_Source001 AS Air_Source
Column001 AS Column
Material_source001 AS Material_source
Fuel_source001 AS Fuel
Secondary_source001 AS Secondary
TOPOLOGY
Material_source001.Material = Column001.Material_Inlet;
Column001.Air_Inlet = Air_Source001.Air;
Column001.Fuel_Inlet=Fuel_source001.Fuel;
Column001.Secondary_Inlet=Secondary_source001.Secondary;
A.8 Process Entity Firing Column
UNIT
Tile AS Firing_column
ASSIGN
# Start Unit Specifications
WITHIN Tile DO
WITHIN Air_Source001 DO
AirMassflowrate := 0.5 ;
AirMassFraction(1) := 0 ;
AirMassFraction(2) := 0 ;
AirMassFraction(3) := 0.21 ;
AirMassFraction(4) := 0.79 ;
AirMassFraction(5) := 0 ;
AirTemprature := 800 ;
PressureInlet := 100000.0 ;
END # WITHIN Air_Source001
WITHIN Column001 DO
WITHIN Material_Source001 DO
Temperature_car := 550.0 ;
Temperature_tiles := 550.0 ;
END # WITHIN Material_Source001
END # WITHIN Column001
WITHIN Fuel_source001 DO
FuelMassflowrate := 0.005 ;
FuelMassFraction(1) := 0 ;
FuelMassFraction(2) := 0 ;
FuelMassFraction(3) := 0 ;
FuelMassFraction(4) := 0 ;
Imperial College London Arnaud
Schalk
2010
72
FuelMassFraction(5) := 1 ;
FuelTemperature := 298.15 ;
PfuelInlet := 100000.0 ;
END # WITHIN Fuel_source001
WITHIN Material_source001 DO
Temperature_car := 516 ;
Temperature_tiles := 1000 ;
END # WITHIN Material_source001
WITHIN Secondary_source001 DO
PsecInlet := 100000.0 ;
SecMassflowrate := 0.5 ;
SecMassFraction(1) := 0 ;
SecMassFraction(2) := 0 ;
SecMassFraction(3) := 0.21 ;
SecMassFraction(4) := 0.79 ;
SecMassFraction(5) := 0 ;
SecTemperature := 700 ;
END # WITHIN Secondary_source001
END # WITHIN Tile
# End Unit Specifications
SCHEDULE
SEQUENCE
CONTINUE FOR 2000
End
Imperial College London Arnaud
Schalk
2010
73
A.9 Model Firing Column Topology