A predictive kinetic model of sulfur release from coal T. Maffei, S. Sommariva, E. Ranzi, T. Faravelli ⇑ Dipartimento di Chimica, Materiali e Ingegneria Chimica, Politecnico di Milano, Piazza Leonardo Da Vinci 32, 20133 Milano, Italy article info Article history: Received 17 November 2010 Received in revised form 28 July 2011 Accepted 2 August 2011 Available online 19 August 2011 Keywords: Sulfur Coal Devolatilization Modeling Kinetics abstract The aim of this paper is to present a predictive kinetic model of the release of sulfur compounds from coal. The model first characterizes the relative amount of the different forms of sulfur components. Inor- ganic (pyritic and sulfates) and organic sulfur (aliphatic, aromatic and thiophenic) are estimated quite simply on the basis of a wide range of elemental coal compositions available in the literature. Sulfur com- pounds are released in the gas phase mainly as H 2 S and S components in the tar fraction. A sulfur fraction also remains in the residual char. The kinetic model, with the related rate parameters, is validated through comparison with experimental data from the literature. The agreement is satisfactory, even though further experimental and theoretical investigation would be useful to improve the reliability of this kinetic model still further. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction Coal is becoming an increasingly important energy resource in the modern world. Its relatively low cost, widespread availability and distribution, plus the fact that it is less vulnerable to political constraints, make it the most attractive fuel for the electricity pro- duction, particularly in developing countries. That said, coal is a cause of environmental concern and not only because of the greenhouse effects resulting from the CO 2 it emits. To make coal more attractive than other fossil fuels, new more effective and environmentally sustainable technologies need to be developed. Fundamental new kinetic studies and research are required to improve our understanding of the multiscale and multiphase phenomena occurring during coal gasification and combustion. With their negative impact on the environment and human health, sulfur oxides (SO x ) are significant pollutants created during coal combustion. Sulfur content in coals is generally in the 0.5–2 wt.% range, but can go up to and above 10% [1]. The release of sulfur species in the gas phase during coal devolatilization is responsible for successive SO x formation, thus its characterization is the first crucial step in monitoring this pollutant emission. The release of sulfur compounds, parallel to coal devolatilization, is the result of a complex process, which involves many interactions between chemical and physical phenomena. The rank and proper- ties of the coal as well as the nature and amount of sulfur involved significantly influence heat and mass transfer as well as reaction rates. Therefore, reaction times, yields, and emissions all depend on the original source [2]. Coal pyrolysis releases sulfur as gas species (H 2 S, COS, SO 2 , CS 2 ) and mercaptans in the tar phase, while the rest remains in the solid matrix of the residual char. Generally, H 2 S is the most abundant gas component, and there is usually a significant amount of mer- captans too [3,4]. In other cases [5,6], large amounts of SO 2 are also revealed. Typical kinetic models of sulfur release from coal pyrolysis refer to empirical models, which define kinetic parameters on the basis of ‘ad hoc’ experimental data with specific coals and reaction con- ditions. The release of sulfur species with the one step model [4,6,7] is described simply as: dV XS dt ¼ k 0 exp  E RT   ½V  XS  V XS  ð1Þ where V XS is the sulfur fraction released and V  XS is its maximum va- lue. The frequency factor, activation energy and released fraction are estimated on the basis of experimental measurements and are greatly dependent both on the original coal and the experimental conditions. The distributed activation energy models [5] overcome some difficulties and are better able to characterize the devolatilization process across a wider range of conditions. The activation energy here is assumed with a density probability function, typically a Gaussian one: V  XS  V XS V  XS ¼ Z 1 0 e  R t 0 k 0 e ð E RT Þ dt h i  1 r ffiffiffiffiffiffi 2p p e  ðE EÞ 2 2r 2 h i dE ð2Þ 0016-2361/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2011.08.017 ⇑ Corresponding author. Tel.: +39 022399 3282. E-mail address: tiziano.faravelli@polimi.it (T. Faravelli). Fuel 91 (2012) 213–223 Contents lists available at SciVerse ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel