Reducing the SO2 emission from a cement kiln

Previous Post
Next Post


Reducing the SO2 emission from a cement kiln



Sayed Horkoss*
Kaferaka,El Koura, Lebanon; Fax No.009616545088. ; Tel No.009613381527; E-mail:


The use of fuel containing high sulfur is becoming one of the few choices for the cement industry due to:

  • The limited quantity of fuel on our planet according to conducted investigations and due to the dramatic cost increase of fuel;
  • The high amount of CaO neutralizing the SO2 in the gas and assuring the desulphurization;

The stabilization of sulfur compounds in the end product (Clinker) will be affected by many factors such as:

  • The burning zone temperature and burner type;
  • The percentage of oxygen in the kiln system and under the flame;
  • The gas velocity inside the kiln tube;
  • The retention time of the clinker at high temperature;
  • The amount of alkalis (Na2O & K2O) in the raw material;

And many other factors affecting positively or negatively the reduction of emissions.

The influence of the burning zone temperature is indirectly determined by monitoring the burnability of the raw mix. The aim of our study is to reduce the SO2 emission without affecting the quality of the end product. For this reason we regulated the mineralogical composition of the raw material without changing its chemical composition.

We found out that the reduction of the burnability index of the raw mix to less than 60 (Polysius method) reduces the emission of the SO2.

Furthermore, reducing the sulfide in raw material to less than 0.20 % by substituting the source of material by other material containing less percentage of sulfides will give good results.


Nowadays, the limited availability and the high cost of fuel start to be a problem for the industry using high amounts of energy in the production process such as cement companies. Because of its relatively low cost, high sulfur fuel is currently one of the best fuel choices for cement industries.

However, increasing the input of sulfur in the kiln system s from the fuel may increase the emission of SO2 from the main chimney. This will lead to many environmental problems such as acid rain which has several environmental and health drawbacks like the acidification of surface water and soil, the reduction of biodiversity and the deterioration of human health . Therefore, the use of fuel with low sulfur content is preferable.

Fuel substitution in a particular cement plant may not be economically viable because energy cost nominally represent about one third of the cost of cement manufacture.

Consequently the reduction of sulfur in the kiln feed would be more feasible solution.

The aim of our study is to reduce the SO2 emission by changing the mineralogical composition of the raw materials without affecting the quality of the end product.

Before exposing the results of our work,we will start by reviewing the basic principles related to cement production and SO2 emission and studying the essential parameters influencing a reduced SO2 emission from the chimney.

Following this literature review,we will discuss the methodology we have to adapted in order to reduce the SO2 emissions.

Finally, we will expose results of our work in this regards namely the importance of working with raw materials having a reduced sulfur content and low burnability index in order to lower the SO2 levels in the effluent gas.


Cement manufacturing is an energy intensive process; about 80% of the total energy required in the cement production is consumed in the thermal conversion of raw feed into clinker.

Typically a wet kiln consumes 1400kcal/kg clinker produced compared to a modern kiln (with preheater and precalciner) which consumes only 700kcal/kg clinker produced (BHATTY et al., 2004).

The final choice of a fuel is always a result of a match among three factors that can be shown, as an equilateral triangle – namely, cost, product quality, and the impact on the environment .The fuel that best meets the specific needs of a cement plant is that which is ideally placed “closest” to the centre of gravity of this triangle (BHATTY et al., 2004).

Based on this triangle many cement producers found that, the pet coke is one of the best choices.

Additionally according to the “Circulaire du 19 juin 1985, journal officiel du 7 août 1985” from the French Environmental Ministry the pet coke was accepted as fuel for the cement industry (CIRCULAIRE, 1985).

Pet coke is a black, shiny solid presented as small granules or needles. It is an oil refinery byproduct that results from the thermal decomposition of heavy oils. In the composition of pet coke, we find mainly carbon although it can also show high levels of sulfur (BHATTY et al., 2004).

Chemical reactions in classical operation

The main chemical reaction in the kiln operation process is the transformation of the calcium oxide and silicate oxide to calcium silicates (partly C2S or Ca2SiO4 and mainly C3S or Ca3SiO4).The temperature rise in the kiln and the quick cooling of the clinker will have a big influence on the stability, the reactivity, the crystal size and the distribution of the new phases of the clinker. Both parameters will affect the strength and setting time of the cement produced.

Possible Reactions in the presence of high sulfur

The sulfur is introduced into the kiln system from the two mains feeding sources: The fuel and the raw materials (HORKOSS, 2004). From the fuel: In the main burner flame of the kiln and in the precalciner burner, all the sulfur in the fuel will burn to SO2 according to the following reaction: S + O2 → SO2

From the raw material: The sulfide and sulfur combined in organic compounds burn to SO2 in the upper stages of the pre-heater: Organic S + O2 → SO2 and Sulfide +O2 → Oxide + SO2

Also some raw materials contain traces of gypsum as CaSO4, CaSO4 2H2O, CaSO4 1/2H2O which decompose or react with other components at high temperature.

The SO2 integrated in the system from both sources will react with the CaO, K2O, and Na2O present in the raw mix and with the intermediary and final crystals of the clinker or leave the kiln as emission SO2 from the main chimney.

The reactions of SO2 with the alkalis (Na2O and K2O) present in the raw mix are mainly homogeneous reactions as they take place in the gas phase. Reactions of the flue gas containing SO2 with calcium compounds present in the raw mix, are heterogeneous reactions (i.e. reactions between two phases).In a heterogeneous reaction the first step is adsorption (the adsorbed quantity depends on the temperature and surface area). The second step is diffusion.

Laboratory tests showed a maximum SO2 reaction with CaO at approximately 800°C (KREFT and SCHÜTTE,1983).
Some of the possible reactions of SO2 during the clinker production process:

CaCO3 + SO2 → CaSO3 + CO2 This reaction will start by the adsorption of the SO2 in the raw mill (MORTENSEN et al, 1989)

  • CaO + SO2 + ½ O2 → CaSO4
  • K2O + SO2 +½ O2 → K2SO4
  • Na2O+ SO2 +½ O2 → Na2SO4
  • 2CaSO4 + K2SO4 → Ca2K2(SO4)3 or called Ca (BHATTY,1996;TAYLOR ,1998; TASOUMELEAS, 2000 ; HORKOSS, 2004)
  • The reaction of the SO3 with the solid or liquid phase in the kiln tube produces many new phases such as [Ca5(SiO2)2(SO4)] and [Ca(Al6O12)(SO3) ] (TAYLOR, 1998).

Additionally to the above two outputs,there is one undesirable possibility that some of the sulfur is retained in the kiln tube or system as deposits (BHATTY et al., 2004), because it will lead to a blockage in the kiln system.

Cycle of Sulfur

Most of cement plants operate under oxidizing conditions with slight excess of oxygen (0.8 to 2%) in the kiln inlet The continuous excess of oxygen in the combustion atmosphere ensures that sulfur, even if introduced chemically in reduced forms (S1-, S2-), will be oxidized to S4+ and S6+.This why the percentage of oxygen needed will depend on the form and percentage of sulfur (BHATTY et al., 2004).

Under an oxidizing atmosphere,the sulfur of the fuel will be oxidized in the kiln burner(kiln outlet) to SO2 ,which at lower temperature (kiln inlet) will react with the free lime to produce anhydrite CaSO4 . The presence of the new element (CaSO4) within the kiln feed in the kiln tube may improve the feed burnability by contributing to low temperature melt formation, where it lowers melt viscosity and promotes solid state and fluxed reactions in the lower temperature zone of the kiln(VISWANATHAN and GHOSH, 1983).

When the calcium sulfate and alkali sulfate reach the kiln outlet (temperature around 1500°C) part of this material will be decomposed while other lives the kiln with the clinker. This phenomenon will increase the concentration of SO2 in the kiln atmosphere. When the SO2 concentration in the kiln atmosphere reaches certain value (depending on the condition of each kiln) it will live the kiln tube as gas emission .

Available methods for monitoring the SO2 emission

According to CEMBUREAU (CEMBUREAU, 1999), the SO2 emissions of cement kilns vary between 10 and 3500 mg/Nm3. To reduce the SO2 emission different methods are available:

1- Reduction of Sulfur contents

This method is based on the substitution strategy aiming at reducing the input in order to reduce the output; in general it is the easer way to reduce the emission.But because energy costs nominally represent about one – third of the cost of cement manufacture, fuel substitution at a particular plant may not be economically viabl .Cement plants are normally located at or near their sources of raw materials, and there are often critical economic limitations to the practicability of substituted raw materials to reduce the input of sulfide sulfur.

2- Optimization of the clinker burning process

Optimization of the clinker burning process is usually done to reduce the heat consumption,to improve clinker quality and to increase the life time of the equipment. Thereduction of SO2 emission is only a side effect of the optimization.
This method needs the application of a high level control system and it is affected by many parameters related to the fuel, raw material, the burner and flame shape. For these reasons, each plant needs a special implementation and additional technology.

3- Addition of Ca(OH)2 to preheater upper stage.

To serve as SO2 absorbing reagent, hydrated lime can be introduced into an appropriate location in the upper stages of the preaheater tower.The hydrated lime will react with the SO2 directly (Ca(OH)2 + SO2 → CaSO3 +H2O) or the hydrated lime is converted to calcium oxide CaO at 522°C to form an effective scrubbing reagent at the location in the process when sulfide is being converted to SO2 (CaO+ SO2 + ½ O2 → CaSO4). The results of this method are not always acceptable especially for high initial SO2 emission. Under these conditions, the addition of lime is economically and ecologically not feasible.

4- Circulating fluidized bed absorber CFBA

To reduce very high SO2 emissions, ranging from more than 1500mg/Nm3 (normal m3) to 500 or 400 mg/Nm3,a separate scrubber is required if the primary reduction measures remain insufficient .The CFBA uses a venturi reactor column to produce a fluidized bed consisting of a blend of slaked lime and raw meal. The intensive contact between gas and absorbent, the long residence time and the low operating temperature close to the dew point allow a very efficient absorption of SO2. The gas leaving the venturi is loaded with absorbent which is collected in the downstream electrostatic precipitator. The maximum SO2 reduction efficiency is not a guaranteed efficiency for the application on any kiln but the maximum efficiency that may be achieved under optimum conditions. The maximum reduction efficiency is based on an initial SO2 emission of about 3000mg/Nm3 which means that only in case of very high SO2 emissions, the application of this method shows high efficiency.

5- Wet Scrubber

An alternative to the CFBA (dry scrubbing) is wet scrubbing. The exhaust gas from the kiln passes first a gas/water heat exchanger before it enters the SO2 scrubber at a temperature of about 115 °C. In the scrubber the SO2 is absorbed in slurry loaded with 6 to 10 % solids consisting of 98% CaSO42H2O and 2% CaCO3. The slurry is sprayed in counter current to the exhaust gas and collected in the recycle tank at the bottom of the scrubber where it is oxidized with air (CaSO3 +1/2 O2 → CaSO4 ). A part of the slurry is pumped to a centrifuge where water and gypsum are separated. The rest is re-injected through a circulation line into the scrubber. Chalk slurry of 30% moisture is injected into the circulation line before the spray nozzles to replace the used and extracted absorbent (CaCO3 + SO2 → CaSO3 + CO2). The exhaust gas leaves the scrubber with a temperature of about 70°C. This system is very costly and as in the previous one, the maximum reduction efficiency is based on an initial SO2 emission of about 3000mg/Nm3.

6- Adsorption om activated coke.

The de-dusted kiln exhaust gas is passed across the activated coke where compounds like SO2 and NH3 are efficiently adsorbed. The cleaned gas is then released the atmosphere. The used activated coke is periodically extracted to a separate silo and replaced with fresh adsorbent. Adsorption on activated carbon is too expensive and it is installed only in a Swiss cement plant.

7- Calcined feed recirculation.

F.L. Smidth has developed a proprietary process, De –SOX, in which small quantity of partially calcined feed (e.g. 5%) is removed from the calciner vessel of a pre-calciner kiln system and pneumatically conveyed to an appropriate point in the upper stages of the pre-heater tower. The calcium oxide in the calcined feed is an effective scrubbing reagent at the location in the process where sulfide sulfur is being converted to SO2.

8- Cement Kiln Dust Internal Scrubber

F.L. Smidth has developed another proprietary process, Gas suspended absorption (GSA), in which dry lime rich cement kiln dust from the alkali bypass on a pre-calciner kiln system is re-circulated to the conditioning tower ahead of the bypass in the gas flow path. In the presence of water in the conditioning tower, this calcium oxide becomes an effective SO2 scrubbing reagent.

Factors influencing the SO2 Emission

The heat transfer process occurring inside the kiln system is extremely complicated. Many factors affect the optimization of the kiln and the quality of the clinker.The raw mix characteristics (preparation, homogeneity, chemical composition, mineralogical composition, fineness …) will have a major impact on the kiln stability and production quality. They will affect also the refractory life, fuel economy and environment impact.

The objective of our study is to find a sustainable solution to reduce the SO2 emission from a cement kiln when using high sulfur fuel.Most of the methods that were listed before are either substitution or end pipe solution, for this reason we started to analyze the direct (the sulfur form in the raw materials) and indirect (the stabilization of the sulfur compound in the clinker phases) factors affecting the emission of SO2:

  • Form of sulfur content in the raw material

As described before, the sulfur is introduced into the kiln from raw materials as sulfate such as calcium sulfates, in chemically reduced forming minerals such as pyrite and as organ-sulfides that are constituents of oil or tar. The sulfur that enters the kiln in the raw material in the form of sulfate will tend to be retained within the kiln system; however the sulfur that enters in the form of organic sulfur compounds or sulfides may be emitted as SO2.

  • Burning zone temperature and burner type

The most important mass flows in the clinkering cycle are the counter current flows of solid and combustion gases. The internal transport cycles within the kiln are of considerable importance to mass transport and influence the kiln operation. Sulphur is associated as one of the complex cycles which affect the kiln operation (BHATTY et al., 2004).
The burning zone temperature is the main factor affecting the volatilization and decomposition of the sulfur compounds. A higher burning zone temperature, increase dramatically the decomposition and volatilization and lead to process and environmental problems (CLARK, 2003).
The type of the burner has an indirect action on the stability of the sulphur compound because it will affect the percentage of oxygen, the flame length & shape and the burning zone temperature.
Based on the above, we can summarize the influence of the burner type as follow: optimizing the length and the shape of the burner flame, and increasing the percentage of oxygen under the flame without increasing the velocity in the kiln by using a modern and expensive burner may stabilize the sulphur compounds in the clinker.

  • Burnability of the raw material

The raw materials used are designed and proportioned to provide the appropriate amount of the various clinker phases; burnability is the measurement of the transformation of the belite (C2S) and free lime to alite (C3S) into the desired clinker phases.
The formation of alite, belite and free lime depends mainly on the burning zone temperature and the temperature profile in the kiln which means the easier transformation (low burnability); the lower the burning zone temperature. Therefore the burnability is the most important point for the clinker production (HILLS, 2003). In addition, the stabilization of the sulfur compounds in the clinker is mainly related to the burning zone temperature. The lowest temperature leads to the highest stabilization.
Apart from the temperature there are also direct and indirect factors influencing the raw material burnability:

the chemical composition as a direct factor and the mineralogical composition as an indirect factor.

a – Chemical composition of the raw materials

Theoretically, the chemical composition is the main factor related to more or less free lime which leads to hard or easy burnability:

  • The lime saturation factor LSF represents the CaO of the mix; a higher LSF leads to harder burnability and vice versa (Möller, 1997).
  • The silicate ratio is related to the amount of liquid phase at the burning zone temperature. More liquid phase improves the burnability (Möller, 1997).
  • Flux such as high iron or fluoride improves the burnability of the raw mix (Möller, 1997).
  • Mineralizes such as sulfur also improve the burnability (i.e. lowers the viscosity of the clinker melt allowing quicker reaction velocity)(Möller, 1997).

B -Mineralogical composition of the raw materials.

This factor can affect the raw mix burnability more than the chemical composition for many reasons:

  • The atoms have to travel certain distances in order to meet and perform the chemical reaction so it is important to have well homogenized raw material with suitable fineness (Möller, 1997).
  • The fineness of the raw meal is a very important factor up to a certain level. Too coarse raw meal hinders the fusion of the atoms; too fine particles disturb the material flow through the kiln and preheater leads to blockage, rings and liquid clinker (Möller, 1997).
  • The diffusion of silicon into CaO is four to five times slower than the reverse. Therefore the form of the silicon in the raw mix is very important. Any changes in the silicon form will affect the burnability. When the source of silica contains quartz sand with less reactivity, the burnability will be very hard. When the source of silicate is clay or other type of reactive silicate, the burnability will be improved (Möller, 1997).
  • Too large calcite crystals (CaCO3) in the raw mix will affect the burnability (Möller, 1997).
  • The lower the temperature at which less free lime can be obtained, the better is the burnability of the raw mix (Möller, 1997).
  • If the first belite crystals have been formed very early and have grown too large, the conversion of belite into alite might be too slow or may even be prevented (Möller, 1997).

Based on the above, we can understand the big influence of the raw material burnability on the stabilization of the sulfur compound in the clinker which lead to reducing the emission of the SO2.

  • Percentage of oxygen

The dissociation reaction of the CaSO4 on the burning zone is as follows: CaSO4 → CaO + SO2 +1/2 O2 According to the law of mass action from Goldberg and Waage, the equilibrium constant K based on the concentrations of each reactant can be determined by: (LENE, 2001) K = [CaO] x [SO2] x [O2]1/2 / [CaSO4] Since K is constant, any variation in concentration of one of the compounds will affect the degree of decomposition of the reaction; higher oxygen level will affect positively the stability of the CaSO4 and reduce the generation of SO2 in the kiln system.
The volatilization of the K2SO4 and Na2SO4 will not be affected much but the high level of oxygen can encourage the formation of calcium-langbeinite (reaction between K2SO4 and CaSO4) which reduce the volatilization of K2SO4 and the dissociation of sulfur from the end product (BHATTY et al., 2004).
The kiln flame gets the oxygen from two sources: the primary air and secondary air. The primary air is relatively cold air, with maximum10 percent from the total quantity.The secondary air, originating from the kiln cooler is relatively hot air 800 to 1000°C, and is the main source of air representing more than 90% of the total air.
The control of the percentage of oxygen in the flame is not possible directly so the amount of oxygen in kiln inlet will relatively indicate the level of oxygen in the flame.
Increasing the oxygen content from the primary air will increase the level of cold air under the flame which will affect the ignition of the fuel. Increasing the oxygen content from the secondary air (hot air, better for burning) means increase the preheater fane speed which leads to increasing the velocity in the kiln tube and end to an undesired result.

One of the best solutions is to force the hot oxygen in the secondary air to reach the flame but this method needs a very expensive burner.The monitoring of all this solution to increase the percentage of oxygen needs a very complicated procedure and a full automatic system for quality control which is not available yet.

  • The Gas Velocity

The gas velocity in the kiln tube and burning zone affect the stabilization of sulfur compound in the clinker. A high gas velocity reduces the vapor pressure of the sulfur volatiles in the atmosphere and increases their volatility (BHATTY et al., 2004).

  • The percentage of SO2 in the burning zone

The percentage of SO2 in the burning zone will reduce the volatility of the potassium sulfate (BHATTY et al., 2004).This factor will be achieved immediately when using high sulfur fuel.

  • Time of heating at higher temperature

Reducing the time of heated at high temperature will increase the stability of the sulfur compounds in the clinker.

Orientation of the project Study

Taking into consideration the above factors, the possible solution of the sulfur compounds volatilization and dissociation is first to operate a kiln with an oxidizing atmosphere in the burning zone without increasing the gas velocity in the kiln tube, second by reducing the burning time of the materials at high temperature, third by lowering the burning zone temperature.

Increasing of oxygen concentration is very complicated. Additionally, reducing the burning zone temperature is not possible except if the material need less energy, because this factor will affect the quality of the end product.

The burner type can be helpful, but it is very expensive project.

The time of heating at higher temperature and the gas velocity are related to the kiln operation and the modification is very limited.

Consequently, the only feasible factor to reduce the volatilization and decomposition of the sulfur compounds in the clinker and indirectly reduce the SO2 emission is to lower the Burnability index of the raw mix, and this what will be investigate in the following sections.

Results and discussion

Our goal was to find if we can improve the material burnability without major changes in the chemistry of the raw feed. After performing tests on different raw meals (Table -1), the correlation between the chemical compounds and the raw material burnability was not clear.

Contrary to our expectations we found hard to burn raw meal with a low lime saturation factor LSF (stock S 20) and easy to burn raw meal with high LSF (stock S 9) provided that all the tested samples are from the same raw mill with the same fineness (Table -1) To clarify the previous inexplicable phenomenon we started looking to other factors in the raw material, our new goal was to find the reason of this contradiction between the theory and our finding.

The raw materials at Cimenterie Nationale were divided to six classes (tables -2).

After analyzing the chemical composition of the different raw materials (tables -2), we started some laboratory tests using the six different raw materials from the quarry.

For that, we prepared four mixtures with the same chemical composition but we changed the types of raw materials (Table -3).

The chemical composition of all four mixes (Table -3) was calculated with LSF 99.00; Silicate module S.M. = 2.4 and aluminum module A.M. =1.25 which means that the chemical composition of the four mixtures was the same. We changed only the source of limestone. The residue over 0.09 mm and 0.2 mm sieves were equivalent.

The results in table -4 showed that with the same chemical analysis, same fineness and at the same temperature we got different results of C3S and free lime.

The above listed test results (Table-4) reveal the dependence of the burnability, the Alite formation and the free lime content on the hardness of the limestone, on the place of origin of the limestone and on define proportions in the mix .

To improve kiln operation, it is therefore very important to achieve always a raw mix similar to mixture 3 (Table-4) without major changes.This will enhance the clinker / cement quality, reduce the heat consumption, increase the life time of the bricks and probably reduce the SO2 emission.

We started the implementation of the proposed solution. The first stock S75 (Table-5) show a burnability index of 100 with an average of SO2 emission equal to 230 mg/Nm3 that is acceptable according to German standards VDI 2094 (less than 400 mg/Nm3) (VDI 2094, 2003)while.

After the implementation of the proposed solution, the results of the stocks S76, S77 were better. The burnability index was decreased to around 88 and SO2 emission to around 200 mg/Nm3 without changing the raw feed chemistry or affecting the quality of the end product.

We came back to the existing situation (burnability index 100) for the Stock S78. The SO2 emission was increased to an average of 233 mg/Nm3(Table-5).It was obvious that we got the first proof on the influence of the raw mix burnability on the SO2 emission.

When we reduced again the burnability index of the stocks S83, S84, S85 and S86, the SO2 emission decreased dramatically (Table-5).

After evaluating the test results (Table -5) it was clears that the raw mix burnability had an influence on the SO2 emission. When the burnability index was reduced from 100 to 63, the emission of SO2 was decreased from 230 mg/Nm3 to 81.5 mg/Nm3.But also it is shown that the form of sulfur especially the sulfide in the raw material and not the total sulfur content had also an influence on the SO2 emission. When we compare on one hand the stocks S 76 & S 77 (burnability index 88 and 89) and on other hand the stock S 84 (burnability index 87), we find a big difference in the SO2 emission from around 200 mg/Nm3 to 139mg /Nm3 even with theoretically the same burnability index.

This is due to a lower percentage of the sulfide in the raw material for the stock S84 compared to stocks S76 and S77 .In fact, as it was described before, the sulfide can be in a very easy volatile form and dissociated during the milling or at relatively low temperature. The sulfide in the stocks S 76 & S 77 was 0.26 & 0.25 % but it was 0.21% in the stock S 84 .


As a conclusion reducing the burnability index of the raw Mix to less than 60 (Polysius method) by monitoring the mineralogical composition of the raw material without changing the chemical composition of the raw mix or affecting the quality of the end product (Cement) reduce the emission of the SO2 . Additionally reducing the sulfide in raw material to less than 0.20 % by substituting the source of material by other material containing less percentage of sulfides will give also good results.

Notes and references

3 CEMBUREAU 1999. Best Available Techniques for Cement Industry.





Previous Post
Next Post

Leave a Reply