Umeå University's logo

A tool has been developed that enables prediction of the chemistry in cement production with thermodynamic phase equilibrium calculations. Reactions in gas, solid and liquid phases are calculated in the process from preheating tower, including exhaust gas cleaning, through rotary kiln, clinker cooler and ends at the output of clinker. The simulated values are compared to measured or calculated data from a full scale plant. This is a cement plant producing 2000 t clinker per day using both traditional and alternative fuels. The chemistry model shows good agreement especially on material chemistry at various places in the process and on composition of the clinker. A new way to define fuels is used and is straightforward and reliable. In the future work the model has to be improved and more elements are to be added to the thermodynamic database.

A thermodynamic process model is used as an evaluation tool. Full oxy-fuel combustion is evaluated for circulation of 20–80% of flue gases to the burn zone of a rotary kiln. The full oxy-fuel combustion simulations exhibit altered temperature profiles for the process. With 60% recirculation of flue gases, the temperature in the burn zone is comparable to the reference temperature, and carbon dioxide concentration in the flue gases increases from 33 to 76%. If water is excluded, carbon dioxide concentration is 90%. The partial oxy-fuel combustion method is evaluated for 20 and 40% recirculation of flue gases from one cyclone string to both calciners. Fuel and oxygen feed to the burning zone and calciners are optimised for the partial oxy-fuel scenario. The lowest specific energy consumption is desired while maximising the amount of carbon dioxide theoretically possible to capture. By introducing partial oxy-fuel combustion with 20% recirculation of flue gases in the carbon dioxide string, total carbon dioxide emissions increases by 4%, with 84% possible to capture. Within the limits of the model, the introduction of full oxy-fuel and partial oxyfuel combustion is possible while maintaining product quality. When simulating partial oxy-fuel combustion, the energy consumption will increase even when no power consumption for the production of oxygen is included.

This paper describes the formation of a phosphorous belite solid solution and its impact on alite formation. A sub-solidus phase relation for the ternary system silicon dioxide–calcium oxide–phosphorus pentoxide (SiO₂–CaO–P₂O₅) is reported. The ternary system is based on Rietveld refinements of X-ray diffraction patterns from experimental tests. The overall picture is based on known phase diagrams, relevant Rietveld refinements models, stoichiometric relationships as a function of increasing phosphorus pentoxide concentration and vacancy theories for solid solutions of phosphate belites. A tool is developed for predicting the chemistry of the product as well as the chemistry during heating when producing Portland cement clinker. A thermodynamic database for phase chemistry calculations of clinkering reactions has been created and evaluated. Suitable compounds and solution species have been selected from the thermochemical database included in FactSage software. Some solution compositions have been uniquely designed to allow for the proper prediction of the cement clinker chemistry. The calculated results from the developed database for heating raw materials in cement clinker production and cooling of the product are presented in this paper. The calculated results provide a good prediction of the phases and quantities formed during heating and non-equilibrium cooling. The prediction of the amounts of alite, belite and aluminoferrite phases in the product according to the Scheil method is good. The temperature interval for the existence of all of the major phases is relevant. The thermodynamic data for a solution phase of alite with substituting ions of primarily magnesium oxide and phosphorus pentoxide would improve the predictability of the developed database.

Natural limestones as raw material for OPC clinker manufacturing contribute to emissions of CO2gases during the production of clinker. In addition, the mining of limestone can regionally be controlledby restrictions due to environmental concerns. Slags from the steel industry can replace limestone tominimize the use of the mineral deposits. Both materials have similar chemistry and are compatible asraw materials.Utilizing slags raises questions about how slag particles will react with other raw meal components asthe temperature in the kiln increases during clinker formation. This study establishes the chemical andmineralogical aspects of replacing a portion of the limestone with slags. Of interest is how the materialsreact during the formation of the liquid phase and the formation of phases containing MgO.Three different slags were examined, a basic oxygen furnace slag BOF, a crystalline blast-furnace slagand a granulated blast-furnace slag. In the study, the microstructural causes of reactivity, as well asmineral formation in the transition zone between raw meal components, developing liquid phase andslag particles were studied. Heated raw meals were studied using HT-QXRD, QXRD, SEM andthermodynamic modeling to describe the reactions of particles at higher temperatures. The resultsshow that the formation of clinker minerals is strongly influenced by the type and amount of slag. Thus,a careful selection must be done of both composition and quantity of metallurgical slags for naturallimestone replacement in order to maintain clinker quality.

Analytical methods for characterising cement raw meal during heating in different atmospheres were investigated. The effect of replacing limestone with 10 wt% slag on the formation of incipient belite and precursors of the clinker liquid in the temperature range 600–1050°C was quantified using thermogravimetry, X-ray diffraction and equilibrium calculations. The results showed that when calculating the lime saturation factor, slags were favoured to sand, resulting in lower amounts of quartz and C₂S in the samples containing slag than the reference sample. This suggests that silicon dioxide in slag minerals did not react in this temperature range. The multi-component equilibrium results supported the phase formation sequence established. Allowing for the possible kinetic influences the potential solids solutions offered with the software was a valuable asset. The results showed that the effect of using slags to reduce the carbonate and sand content in a raw meal on potential amounts of incipient C₂S was negative. At present, more detailed knowledge is needed regarding how blast-furnace slag and basic oxygen furnace slag contribute to the formation of intermediary compounds such as incipient C₂S, C₃A, C₂F and C₄AF in the solid phase at temperatures over 1050°C and affect the formation of C₃S.