International Conference on Ceramics, Bikaner, India

International Journal of Modern Physics: Conference Series

Vol. 22 (2013) 231–238
 World Scientific Publishing Company
DOI: 10.1142/S2010194513010179

MODELING OF PROCESS PARAMETERS AND ANALYSIS OF EFFECT OF

by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

VARIABLES IN THE DEAD BURNING OF MAGNESITE IN ROTARY KILN

Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com

B. K. CHAKRABARTI
Department of Materials & Metallurgical Engg.
National Institute of Foundry & Forge Technology
Hatia, Ranchi-834003. Jharkhand.
E-mail: bijoy_2000in@yahoo.com

The challenges of improving sintered MgO raw materials with special high temperature properties
demand higher temperature calcination. In the process of dead burning of magnesite (DBM) in
rotary kiln, the heat expenditures were found to be mainly due to dead burning of magnesite, loss
through exhaust gases, loss through kiln shell by radiation & convection, clinker exit etc. The
calcination process is highly energy intensive and involves various interdependent variable
factors. An attempt was made to build a model and a screening design of experiment was
performed with few process variables to identify the greatest effect of variables on the response
quality. The design variables chosen were raw magnesite (RM) feed rate, kiln rotation (RPM), fuel
consumption and burning zone temperature (BZT). The response variables were exit gas
temperature and density of dead burnt magnesite (DBM). A fractional factorial design was used to
keep the number of experimental runs to a minimum. ANOVA and normal plots were used to
evaluate the effects of different variable factors on the sensory/response properties. The
Experimental Design, ANNOVA, Response surface etc. given an insight of dead burning of
magnesite in rotary kiln. This work had enabled us to correlate the BZT, RPM, RM feed with the
exit gas temp and density of magnesite produced. The result opens up an avenue to look into the
optimum region of operations within the ranges of variables considered in order to minimize the
exhaust gas temp. and to maximize the density of the DBM produced..

Keywords: Dead Burning of Magnesite, Rotary kiln, Modeling & DOE

1. Introduction
The ceramic and refractory industries are the big energy consumers 1,2,3 , where the
thermal energy constitutes an important part of the manufacturing costs. Proper
utilization and conversion of energy plays an important role in the energy management of
any system. DBM production in rotary kiln involves consumption of an alarming amount
of energy, where thermal energy constitutes an important part of the manufacturing costs.
The success of the dead burning process of magnesite rests on proper controlling of the
large number of variable factors responsible for the manufacture of DBM. However in
order to visualize optimum out put it is necessary to segregate the important factors
which may be directly and greatly responsible for the success of the process. If
conservation of energy of the process is of concern then obvious step would be to aim at
higher quality of DBM at a lesser energy consumption. An energy campaign of the kiln

231
 232 B. K. Chakrabarti

system shows 4,5,6 that the potential heat loss areas are through kiln shell, chamber exit
gas, clinker exit etc. Due to high exit gas temperature and high volume the heat loss was
assumed to be quite considerable amount. Therefore, exit gas temperature was considered
as one of the response variables. The quality of DBM may be visualized in terms of its
sintering, grain growth, crystalline size, Sp.gravity etc. In the present work the specific
gravity had been chosen as the other response variable.
Multivariate Data Analysis refers 7,8,9 to statistical technique used to analyze data that
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

arises from more than one variable. This essentially models in reality where each
situation, product, or decision involves more than a single variable. Despite the quantum
of data available, the ability to obtain a clear picture of what is going on and make
Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com

intelligent decisions is a challenge. When available information is stored in database

tables containing rows and columns, Multivariate Analysis can be used to process the
information in a meaningful fashion.
With the use of process data we can predict various effects of variables on responses
and also define optimized solutions to the process. Factorial design is an approach to
perform any experiment for analysis. DOE & ANNOVA, are various techniques used to
analyze multivariate data and depict response patterns.
We had used UNSCRAMBLER-9 software package of CAMO Softwares Ltd. to
create factorial design and analyze effect of variables on responses.
Experimental design is aimed precisely at solving the problems. in a situation where
experiments are costly and, determines which variables are important & provides more
information, and thus will have to be performed with higher priority. The advantage of
designing experiments is that one gains control over data distribution, and ensures further
analysis.

2. Data Acquisition & Kiln Description

2.1. Description of the kiln system and data acquisition

Process : Dry process.
Type : Rotary.
Kiln Length : 84 metres.
Kiln Diameter : O.D = 2.575
Inclination : 3%
Fuel Type : LSHS (Low sulfur heavy stock).
Calorific Value of the Fuel : 10800 k.cals / Kg.
Fuel / DBM : 0.185.
Average Exit gas temperature : 2 350C.
Atmospheric temperature : 300C.

Data acquisition and utilization of the data needs proper attention, since there are certain
variable factors which influences the results very much.
 Modeling of Process Parameters in the Dead Burning of Magnesite in Rotary Kiln 233

3. Experimental
There are many variable factors involved in dead burning of magnesite in a rotary kiln..
However, based on data acquisition and the work as reported 4,5,6 the variable factors
chosen in this work were Raw magnesite (RM) feed rate, kiln rotation (RPM) and
burning zone temperature (BZT) and estimated their effect on two responses viz,
Sp.gravity of DBM and exit gas temperature. Sp.gravity of magnesia influences the
quality of product and exit gas temperature influences the energy efficiency of the
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

process. UNSCRAMBLER-9 statistical software package was used to design our

experiment. Experimental design was made with two levels (upper and lower limits) of
Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com

the input variables were performed with two centre sample.

Table 1: Lower and Upper limit of the input variables.

Variables Lower limit Upper limit

Raw material feed 10080 10320
Specific rotational speed 60 80
Burning zone temperature 1550 1650

Table 2: The Data table of Design matrix.

RM feed (kg/hr) RPM (seconds) BZT(°c) Exit temp. Sp.gravity

Run [A] [B] [C] (°c) (g/cc)
1 10080 60 1550 230.0 3.250
2 10320 60 1550 220.0 3.200
3 10080 80 1550 228.0 3.350
4 10320 80 1550 215.0 3.300
5 10080 60 1650 235.0 3.450
6 10320 60 1650 240.0 3.350
7 10080 80 1650 240.0 3.452
8 10320 80 1650 235.0 3.380
9 10200 70 1600 230.5 3.332
10 10200 70 1600 231.5 3.328

Table 3: Analysis of Effect shows Significant/Non significant Effect

Variables Exit Gas Temp Sp.gravity

RM Feed(A) (-)Ve NS
SPR(B) NS NS
BZ Temp(C) + ++
 234 B. K. Chakrabarti

Table 4: ANOVA Table where Y-var is Exit Temp.

Summary SS DF MS F-ratio p-value

Model 437.500 4 109.375 15.670 0.0049
Error 34.900 6.980
Adjusted Total 472.400 9.0 52.489
Intercept 6.03e+05 1 6.03e+05 8.64e+04 0.000
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

RM Feed 253.125 1 253.125 36.264 0.0018

SPR 28.125 1 28.125 4.029 0.1010
Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com

BZTemp. 153.125 1 153.125 21.938 0.0054

Table 5: ANOVA Table where Y-var is Sp. Gravity.

Summary SS DF MS F-ratio p-Value

Model 5.265e-02 4 1.316e-02 21.938 0.0023
Error 3.00e-03 5 6.00e-04
Adjusted total 5.56e-02 9 6.183e-03
Intercept 114.58 1 114.58 1.910e+05 0.000
RM Feed 1.512e-03 1 1.512e-03 2.521 0.1732
SPR 1.250e-05 1 1.250e-05 2.083e-02 0.8908
BZTemp 4.961e-02 1 4.961e-02 82.688 0.0003

Fig. 1: Response Surface Plot:

Y-var: Exit Temp
 Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

Y-Var: Sp.Gravity
Fig. 2: Response Surface Plot:

Fig. 3: Main Effects Plot RM Feed vs Exit Gas Temp.

Modeling of Process Parameters in the Dead Burning of Magnesite in Rotary Kiln
235
 Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

236
B. K. Chakrabarti

Fig. 4: Main Effects Plot of Kiln Rotation vs Exit Gas Temp

Fig. 5: Main Effects Plot of Burning Zone Temperature vs Exit Gas Temp
 Modeling of Process Parameters in the Dead Burning of Magnesite in Rotary Kiln 237
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.
Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com

Fig. 6: Main Effects Plot of Burning Zone Temperature vs Specific Gravity.

4. Results and Discussions

Main effect plot depicts:

• Exit gas temperature decreases as the raw material feed increases as shown in
fig.3. Exit temp has been found to increase as the time of rotation of the kiln
increases as shown in fig.4.
• Exit gas temperature increases with increase in burning zone temperature shown
in fig. 5.
• Fig. 6 illustrates that the specific gravity of DBM increases as the burning zone
temperature increases.

5. Conclusion
There are many process variables for dead burning of magnesite in rotary kiln. However,
due to complexity involved in data acquisition & exeperiments only three variables were
taken into account. Numerous other possibilities and combinations of variables and
responses may be aimed at.
The Experimental Design, ANNOVA & Response surface , main effect plots etc.
given an insight into the interactions among variable factors in the dead burning of
magnesite in rotary kiln. This work had enabled to correlate the BZ. Temp., RPM, RM
feed with the exit gas temp and sp. Gravity of magnesite produced. Table 2. Describes
details of data correlating various input factors and response data. Analysis of main
effect plots fig.3 to fig.6 illustrates that exit gas temperature decreases as the raw
material feed increases and rotational speed increases. The exit gas temperature increases
with increase in burning zone temperature. Main effect plot for Sp. gravity response
shows that Specific gravity increases as the burning zone temperature increases. Results
 238 B. K. Chakrabarti

obtained in the work was an outcome of systematic approach towards data acquisition
and utilization of the same. However, while comparing the results it has to be kept in
mind that the kiln can be judged only after considering the approximations and boundary
conditions which were applied.. A proper monitoring of the kiln is extremely essential
so that the above factors could be regulated properly. This type of work provides data and
appropriate informations to the management of the company about how the units have
been functioning.
by 119.160.69.81 on 05/06/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

References
Int. J. Mod. Phys. Conf. Ser. 2013.22:231-238. Downloaded from www.worldscientific.com

1. John D. Morgan, Mc Kelvey, Ceramics: Energy Update, a report from 79th Annual Meeting of
American Ceram. Society, J. Am. Ceram. Soc.Bull,56 (1977) p-586.
2. Clifford Storey, Energy and the refractories industry, Refractories Journal, 6 (1977)p-26-28.
3. I.K.Smithson, Fuel and energy economy in industry, Tans. British Ceram. Soc.74 (6) 1975, p-
203.
4. “Energy conservation & management in refractory burning units.”- B.K.Chakrabarti, Presented
in the technical session and published in the proceedings of the 65th annual session of the Indian
Ceramic Society, 2002.
5. “Investigations on heat loss through the kiln shell in magnesite dead burning process- a case
study’’- B.K.Chakrabarti, Journal of Applied Thermal Engineering, 22(2002)1339-45.
6. “Studies on the heat transfer behaviour in a magnesite kiln and its assessment”-
B.K.Chakrabarti, Journal of High Temperature Materials & Processes, London, Volm.21, No.4,
2002. P187-194.
7. Montgomery, D. C. “Design and Analysis of Experiments”, Wiley, NY (2004)
8. Dean A. and Voss D. (1999) “DESIGN AND ANALYSIS OF EXPERIMENT” SPRINGER
TEXT in statistics, NEW YORK.
9. T. W. Anderson, “An Introduction to Multivariate Statistical Analysis”, 2nd Edition (New
York: John Wiley & Sons, 1984).