Alternative Fuels in Cement Production

Morten Boberg Larsen

Research output: Book/ReportPh.D. thesis

1710 Downloads (Pure)

Abstract

The substitution of alternative for fossil fuels in cement production has increased significantlyin the last decade. Of these new alternative fuels, solid state fuels presently account for thelargest part, and in particular, meat and bone meal, plastics and tyre derived fuels (TDF)accounted for the most significant alternative fuel energy contributors in the German cementindustry. Solid alternative fuels are typically high in volatile content and they may differsignificantly in physical and chemical properties compared to traditional solid fossil fuels.From the process point of view, considering a modern kiln system for cement production, theuse of alternative fuels mainly influences 1) kiln process stability (may accelerate build up ofblockages preventing gas and/or solids flow), 2) cement clinker quality, 3) emissions, and 4)decreased production capacity. Kiln process stability in particular is influenced by insufficientcarbon burnout in the calciner system, which results in reducing conditions in the materialinlet of the rotary kiln and consequently an increased tendency to form deposits induced bysticky eutectic melts. Clinker quality is mainly affected by minor components from the fuelashes or from carbon dropping into the material charge of the rotary kiln. As regards thepresently most used solid alternative fuels, phosphorous from meat and bone meal or zincfrom TDF are the main components to consider with respect to clinker chemistry. Theemissions seem not to have been affected by the alternative fuels used up until now. However,caution should be taken with regard to emissions of CO when using alternative fuels.As alternative solid fuels are typically high in volatile content, the devolatilization stage in thecombustion process is responsible for a large part of the fuel heating value. In addition, thedevolatilization time of alternative fuels cannot be neglected in kiln system process analyses,as these fuels are typically in the cm-size with devolatilization times in the order of minutes.The devolatilization characteristics of large particles of tyre rubber were investigated in twoexperimental setups with the emphasis being on devolatilization rates and times, and theresults were analysed using mathematical modelling. During devolatilization, the large TDFparticles formed a crackling char layer, which was seen to be removed depending on whetherexternal mechanical interaction was present or not. Both pathways were investigatedexperimentally and a significantly shorter devolatilization time was observed in the situationwhere the char layer was removed. In addition, the experiments showed a significant effect ofparticle size on devolatilization time, where increased particle size increased thedevolatilization time. Model analyses demonstrated that the overall devolatilization kineticsof large particles of tyre rubber is mainly controlled by heat transfer and intrinsic pyrolysiskinetics, whereas mass transfer has negligible influence. The models developed are used topredict devolatilization conversion times for tyre rubber as a function of relevant parameters.The devolatilization rates of other alternative fuels are also expected to be controlled byconversion pathway, heat transfer and intrinsic kinetics.The char combustion stage has a decisive influence on the fuel carbon burnout in cement kilnsystems. The oxidation kinetics of a char from TDF was investigated experimentally and bymathematical modelling. Experiments were performed in a fixed bed reactor under well- iii -defined conditions, where small particles (102-212μm) of TDF were combusted at 750-850°Cat up to 10 vol.% O2. The effluent of the reactor was analysed for CO and CO2, and used toderive conversion against time. The experimental data demonstrated that mass transfer wasimportant within the investigated temperature range of 750-850°C, and a mathematical model for intra-particle diffusion and reaction was developed in order to analyse the data. A reaction expression for the intrinsic kinetics of TDF char oxidation was proposed and comparison with literature data showed fair agreement. For larger TDF char particles with realistic sizes up to 7mm, it was demonstrated that they are converted according to a shrinking particle mechanism, and based on this observation a model was developed in order to explain the controlling factors for TDF char combustion under conditions relevant to cement kiln systems. It was demonstrated that external mass transfer was the rate limiting parameter, as the kinetics are sufficiently faster than external mass transfer. The intrinsic kinetics of other typical alternative fuels is demonstrated to be comparable or faster than intrinsic TDF char oxidation kinetics. Consequently, the char oxidation stage for large particles of other alternative fuels is also expected to be controlled by mass transfer, under conditions relevant to cement kiln systems. A comparison between the mechanisms behind the devolatilization and char combustion stages for large alternative solid fuel particles indicates that the devolatilization kinetics are mainly controlled by heat transfer and intrinsic kinetics, whereas char oxidation kinetics are mainly controlled by mass transfer, under conditions relevant to cement kiln systems. Measurements at two industrial HOTDISC’s using TDF indicate that up to about 75% fuel conversion takes place before discharge into the subsequent calciner. Model analyses of the measurements, using the previously developed sub-models for devolatilization and charoxidation rates, explain that devolatilization takes place in the HOTDISC whereas charoxidation takes place both in the HOTDISC and in the calciner system.
Original languageEnglish
PublisherTechnical University of Denmark, Department of Chemical and Biochemical Engineering
Number of pages222
Publication statusPublished - 2007

Fingerprint

Dive into the research topics of 'Alternative Fuels in Cement Production'. Together they form a unique fingerprint.

Cite this