A Study of Alkali Aerosol Formation in Biomass-coal Co-Combustion with a Mechanistic Approach

A Study of Alkali Aerosol Formation in Biomass-coal Co-Combustion with a Mechanistic Approach
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  10th Conference on Energy for a Clean Environment 7-10 July 2009 A STUDY OF ALKALI AEROSOL FORMATION IN BIOMASS-COAL CO-COMBUSTION WITH A MECHANISTIC APPROACH E. Castellani 1 , M. Falcitelli 1 , L. Tognotti 2  (*) 1. Consorzio Pisa Ricerche –Divisione Energia e Ambiente SCaRL, Lungarno Mediceo 40, 56127 Pisa – ITALY e.castellani@cpr.it, m.falcitelli@cpr.it 2. Università degli Studi di Pisa – Dipartimento di Ingegneria Chimica,Chimica Industriale e Scienza dei Materiali, Via Diotisalvi 2, 56100 Pisa – ITALY tognotti@ing.unipi.it Abstract   A mechanistic approach was developed to simulate inorganic aerosol formation from biomass and coal co-combustion processes. The model encompasses a reactor network model to represent experimental set-up conditions, a detailed kinetic scheme for gas phase reactions and a condensation model for particle nucleation and growth. It was validated by the comparison with experimental data for both the gas phase chemical speciation and size particle distributions. Fine agreement was obtained by using the detailed kinetic scheme with the addiction of a fast irreversible reaction describing the passage of K 2 SO 4  from gaseous to solid phase. The calculations performed with the condensation model suggest how aerosol chemical composition is related to the size distribution. Further efforts in will be devoted to better quantify the relationship between size distribution and chemical composition of aerosol. With more validation and new experimental data the model will be a useful tool for controlling the aerosol formation by the combustion processes. Keywords Inorganic   aerosol, biomass-coal co-combustion, chemical kinetics, reactor model. INTRODUCTION The formation and emission of fine particles from biomass and coal combustion have received increasing attention in the last few years. Aerosols from fuels with high volatile sodium or potassium compounds, as biomass and bio-refuse feedstocks, are often dominated by a large submicron particle fraction, formed as a consequence of alkali volatilization in the combustion zone and the following condensation. The growing interest in these particles mainly owns to their contribution to environmental pollution on both urban and regional scale. This issue is potentially enhanced by the greater difficulties in filtering the ultrafine particulate than the coarser fraction, using the existing flue gas cleaning devices. Moreover, aerosols can be enriched in heavy metals volatilized in the combustion zone, as particularly observed in coal combustion. Fine particles, which in biomass combustion are mainly composed by chlorides and sulphates, also  10th Conference on Energy for a Clean Environment 7-10 July 2009 contribute to deposition (slagging and fouling) and corrosion phenomena on boiler walls and tube surfaces. Alkali reduce the melting temperature of boiler deposits while chlorine accelerates boiler corrosion, which therefore strongly depends on the fuel composition. Nowadays, several studies analyze the role of the gaseous sulphur chemistry on the sulphates/chlorides ratio found in the fine particles from biomass combustion, and suggest the use of co-combustion with fuels containing sulphur (as coal or coke) to minimize corrosion associated with chlorine [1]. The negative effects of aerosol combustion particles in human life and in deposition and corrosion processes of power plants were analyzed in many works [2], even if only very few experimental measurements are available to develop and validate mechanistic models of alkali aerosol formation [1, 3, 4, 5]. From the kinetic point of view, the fundamental reaction scheme of alkali formation is still under investigation. In the published literature there is general agreement in indicating the formation of inorganic aerosol precursors as an homogenous gas phase mechanism, when KCl salts are vaporized in a flue gas stream, simulating coal combustion products [6]. Aerosol gas phase species then condensate to form particles, which grow through coagulation and surface growth processes. Therefore, several authors indicate the necessity of a general model, encompassing both a gas phase and a condensation mechanism. Recently, Glarborg and Marshall [6] proposed a kinetic gas phase mechanism for alkali formation, including an update of the SO x  formation pathways and SO x -alkali reactions. They used the experimental data from Iisa and Lu [3] for validation, showing fine agreement of the predictions with the formation of potassium sulfate under the conditions investigated. Further, the study suggests that a deeper understanding of the condensation and nucleation processes is needed to explain the drain of sulfates out of the gas phase. These preliminary results stimulated the interest in assessing the mechanism for a wider range of reacting environments, mainly those representing industrial operating conditions. Jiménez and Ballester performed a first attempt to apply Glarborg and Marshall kinetic mechanism to real cases of biomass and coal combustion and co-combustion [1]. They simulated the aerosol formation produced in a wide range of test conditions with their experimental setup: an entrained flow reactor (EFR) with a cooling device (called chimeny). The chemical engineering model used in the theoretical study represented only the cooling section of their reactor. They initially used the Glarborg and Marshall complete mechanism, but the calculations produced predictions far away from the experimental data. Then, further simulations have been performed with a simplified mechanism focused only to SO 2 -SO 3  conversion. Even if the comparison of calculated data (Cl/S molar ratio)  10th Conference on Energy for a Clean Environment 7-10 July 2009 with experimental ones showed good agreement for some experimental conditions, their model failed in the low [SO 2 ]-low [O 2 ] region. The past studies showed some difficulties in applying detailed kinetic models for predicting alkali aerosol formation, that should be overcome improving the whole modelling approach. In this work a detailed model for predicting alkali aerosol formation is proposed, based on reactor network analysis [7] and comprehensive mechanistic schemes. One of the novel features is that, besides gaseous combustion species predictions, also particle chemical speciation and granulometric distribution are both included in the calculations. The modeling approach encompasses many sub-models: a reactor network model to represent experimental set-up conditions, a detailed kinetic model for gas phase reactions and a condensation model for particle nucleation and growth. A general combustion mechanism [15] is updated including the entire SO x  and SO x -alkali mechanism by Glarborg and Marshall [6] to simulate aerosol precursors formation. To solve nucleation, surface growth and coagulation problems, the nodal model (NGDE) from Prakash et al. [8] was used, which is essentially an evolution of the sectional model presented by Gelbard [9] and developed in some other works [10, 11]. Finally, the assessment of the modeling approach is achieved, comparing the predictions with the experimental data of aerosol chemical composition and size distribution, measured by Jiménez and Ballester [4]. Since the known technical difficulties to remove fine particles by means of pollution control devices, the final goal is to develop a simulation tool that should be useful to find the best conditions for controlling the aerosol formation in the combustion processes. MODEL DEVELOPMENT The model was developed assuming the alkali aerosol precursor formation as an homogeneous gas phase mechanism. The aerosol particles generate from the inorganic part of the fuel, in the combustion zone, through a process starting with mineral matter vaporization and continuing in the gas phase reactions and finally in particle nucleation, coagulation and growth. Devolatilization is the first step of the process and it strongly depends on the fuel type. Particularly, for what concerns the main alkali aerosol elements (K, S, Cl) from biomass combustion, several studies [5, 12, 13] show that potassium, sulfur and chlorine are almost completely released at 1150°C. The volatiles are then involved in a series of reactions to form aerosol precursors species.  10th Conference on Energy for a Clean Environment 7-10 July 2009 In the present work, a detailed combustion mechanism [15] is combined with Glarborg’s sub-set of reactions for alkali formation, recently published [6]. The gas phase kinetic scheme together with the thermodynamic data of the chemical species and the reactor network model constitute the input program files for the computer software Dsmoke [14], as schematized in Figure 1. The output file provides the gas phase chemical composition of aerosol precursors as a function of the combustion conditions. The general dynamic equation (GDE) was then solved by means of NGDE for the aerosol precursors, obtaining the particle size distribution. One of the aim of this approach is to extend the field of application of the so called Reactor Network Analysis for process study of industrial combustion systems both in pilot and real scale [7]. Figure 1   Computational flow of the modeling approach . Kinetic model description The gas phase reaction mechanism is basically formed on a strongly modular hierarchical structure: the core hydrocarbon combustion has been elaborated by Ranzi et al. [15] and the nitrogen sub-mechanism has been derived by Milan Polytechnic researchers [16] from the work of Miller and Bowman, Kilpinen et al [17] and from the development proposed by Glarborg et al. [18]. The kinetic model was here completed by adding Glarborg’s reactions for the SO 2 → SO 3  conversion and for the alkali species (potassium and sodium sulfates, bisulfates and chlorides) formation. The first step of the Glarborg model [6] is the conversion of SO 2  to SO 3 , which represents the slower and limiting step of the mechanism. This step is mainly based on the following reactions: SO 2  + O(+M) = SO 3 (+M) HOSO 2  + O 2  = SO 3  + HO 2  A minor contribution to the SO 3  formation comes also from: SO 2  + O(+M) = SO 3 (+M) SO 2  + OH= SO 3  + H SO 2  + O 2  = SO 3  + O SO 2  KO(+M) = KSO 3 (+M) Thermodynamic data Kinetic model Kinetic interpreter Kinetic binar file Reactor Network model DSmoke NGDE Particles Size distribution Gas phase composition  10th Conference on Energy for a Clean Environment 7-10 July 2009 KSO 3  + OH = SO 3  + KOH The SO 3  is then converted to K 2 SO 4  by reacting with chloride and potassium hydroxide through the formation of chemical species which are stable at 1573K (KHSO 4  and KSO 3 Cl). Condensation model description The aerosol particle formation depends on both the thermodynamic equilibrium and the transport phenomena. The aerosol system can be described solving the General Dynamic Equation (GDE): cond evapk nuclk coagk k  dt dN dt dN dt dN dt dN   / ... ++=  (1) The NGDE software realized by Prakash et al.[8], solves the GDE by means of a numerical algorithm for nucleation, coagulation and surface growth processes, while several other phenomena are not encountered, such as fragmentation, thermophoresis, diffusion to surfaces and inertial impact. Furthermore, the program has a modular structure able to solve the GDE completely or only for the surface grow or for the nucleation+coagulation process. The software was used by the authors to simulate aluminum particulate formation and growth from a supersaturated phase of the vapor. In the present work NGDE is applied to simulate the particle formation of the main aerosol species, as derived from kinetic gas phase model (i.e. potassium sulfate, potassium chloride and potassium chloride dimer). The Reactor Network Model To build the Reactor Network Model, the experimental setup of Jiménez et al. [4] was represented through a series of interconnected CSTRs (continuous stirred-tank reactors) and PFRs (plug flow reactors) (Figure 2). The R3 to R8 reactors are characterized by a linear temperature profile, to reproduce the temperature profile of the cooling system (Figure 2). The reactors were dimensioned on the basis of the constructive data and operating conditions reported by Jiménez and Ballester [4]. The input gas stream to EFR is composed by natural gas, N 2 , O 2  and SO 2  (stream 1) and a mixture of oxygenate hydrocarbons and alkali species having the same elemental composition of the solid fuel used in the experiments (stream 2).
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks