Elsevier

Fuel

Volume 266, 15 April 2020, 116865
Fuel

Full Length Article
Simulative prediction of ultrafine particulate matter formation by means of different pyrolysis models

https://doi.org/10.1016/j.fuel.2019.116865Get rights and content

Abstract

In this study, the sensitivity of two different pyrolysis models on the simulative prediction of inorganic ultrafine Particulate Matter (PM) formation is investigated. For consideration of PM formation, the volume-conserved discrete-sectional Population Balance Model (PBM) is used. The volatile release is an input parameter of the PBM and thus influences the calculated PM formation. As a representative of an empirical pyrolysis model, the Fu-Zhang model (FZM) is used. In comparison, the Chemical Percolation Devolatilization model (CPD) is taken as a representative for sophisticated pyrolysis network models. The latter takes influences of fuel composition, heating rate and particle temperature into account. The evaluation of the pyrolysis model impact is determined for a pulverized (dp=6574μm) Chinese lignite (Zhundong) since this fuel has a relatively high volatile content and thus favors the formation of particulate matter. To investigate the influence of particle heating rate in the models, two different gas temperatures (1200 K and 1500 K, respectively 5.28·104 K s−1 and 7.51 × 104 K s−1 calculated with the energy balance) are considered. Furthermore, these conditions are selected because experimental data of the PM particle size distribution (PSD) from previous studies are available. The modeling of PSD is performed and validated with the existing experimental PSD data. These experimental data were achieved in investigations with a Hencken burner. The results of the two models (FZM and CPD) show recognizable differences in predicting the PSD. This study indicates that the CPD model improves significantly the prediction of ultrafine PM formation.

Introduction

The formation of inhalable fine particles (dp10μm) is an undesirable side effect of various technical applications [1]. In particular, very fine particles (dp2.5μm) have a high lung damage potential because they have large residence times in the atmosphere and thus can accumulate in the human respiratory system [2], [3], [4]. An additional problem is the concentration of toxic trace elements (e.g. mercury or arsenic) bonded to the ultrafine particulate matter [5]. A significant source of anthropogenic fine particle emissions is the burning of solid fuels like coal or biomass [6], [7]. Coal is still one of the most important primary energy sources for electricity production worldwide [8]. PM emissions from coal-fired power plants in China amount to 3.81 million tons per year [9]. In addition, worldwide coal consumption is expected to increase due to a forecast in electricity consumption, particularly in developing countries [10]. The PM emission may become more severe, by increasing utilization of low-rank coals [11], [12]. Due to the high content of alkali and alkaline earth metals in raw lignite, the emission of PM0.2 can be increased by a factor of five compared to high-rank bituminous coal [12]. The reduction of emissions can either be achieved by a reduction of the primary particle formation or by an increased particle growth coming along with an improved deposition in the filter system [13], [14], [9]. An assessment of these measures requires basic understanding of reaction pathways and relevant parameters.

Coal combustion produces different types of particles with a wide range of sizes, morphologies and components due to different formation mechanisms, e.g. organic particles such as soot and inorganic particles such as ash. Previous numerical studies have developed a basic framework for the formation mechanisms of PM with different size ranges [15], [16], [17], [18]. It is assumed that submicron particles are formed during the entire process of coal combustion by an evaporation-nucleation-coagulation-agglomeration mechanism [19], [20], [21]. The conventional simulation of fine PM formation during coal combustion is mainly based on the redox reaction of minerals as proposed by Sarofim et al. [22], [23]. However, this model predicts the total yield of fine PM (mean values) without temporal and spatial resolution. In addition, detailed, size-resolved information on fine PM cannot be obtained with this model and particle dynamics are neglected. Various physical and chemical mechanisms are involved, such as the evaporation of volatile minerals, the redox reaction of refractory minerals or nucleation, coagulation, agglomeration and sintering [24], [23], [25]. The Population Balance Model (PBM), which is widely used in aerosol science and flame synthesis [26], [27], [28], [29], was used by Gao et al. [30] to simulate fine PM formation in the early stage of lignite combustion. The PBM is based on the General Dynamic Equation (GDE) of particles as formulated in [26]. By correctly defining a collision kernel in laminar flames, to describe the probability of a particle collision, the coagulation process of particles was simulated and the Particle Size Distribution (PSD) of fine PM was predicted. In addition, an important input for the PBM is the amount of minerals released. This release is closely related to the combustion behavior of coal particles in various combustion stages, including devolatilization, char combustion and char burnout [31], [13]. Therefore, an accurate description of coal devolatilization and combustion is relevant for determining the release rate and yield of PM. In the previous work by Gao et al. [30], the Fu Zhang Model (FZM) [32] was applied to determine the volatile release and finally calculate the ultrafine inorganic PM. The accuracy of the pyrolysis model to predict the PM formation has not been investigated by now. Therefore it is necessary to carry out further investigations in this area in order to gain a deeper insight into the PM formation.

The aim of this work is to investigate the influence of the devolatilization description within the PM formation simulation. For this purpose, the PBM is used to describe the PM formation of the early stage devolatilization. Li et al. [12] showed that the ultrafine particles smaller than 0.1 μm are mainly formed by nucleation/condensation of gas-phase minerals. Therefore, our simulations consider only the inorganic particle formation. The Fu Zhang model (empirical model, used in the previous work from Gao et al. [30]) [32] and the Chemical Percolation Devolatilization (CPD) model (sophisticated pyrolysis network model) [33], [34] are applied in this work to calculate the volatile release. To investigate whether the devolatilization model has an impact on predicting PM formation, the calculated PSD data with the two pyrolysis models are compared with experimentally obtained PSD data of a Hencken burner system.

Section snippets

Investigated experimental system

To validate numerical models for the prediction of PM-formation in solid fuel combustion, experimental data are relevant. Therefore Gao et al. [30] has performed experimental investigations with a flat flame (Hencken) burner. The Hencken burner (Fig. 1) enables similar particle heating rates (105 K s−1) as in common pulverized fuel burners [35]. The main advantage of a Hencken burner, however, is that it has a nearly homogeneous radial temperature distribution and a laminar gas flow [35]. This

Modeling

In this section the models to describe PM formation are presented. For this purpose, the description of the coal particle heating rate will be shown, which is decisive for the release of the volatiles and thus also for the PM formation [30]. Furthermore, the reaction mechanism of the ultrafine particulate matter and the volume-conserved discrete-sectional PBM will be presented. Finally, the differences of the pyrolysis models will be described in more detail.

Results and discussion

Fig. 5 shows the comparison of PSDs received from numerical simulations and experiments. The experiments were performed by Gao et al. [30]. For measuring the ultrafine particulate matter, a two-stage isokinetic sampling probe was used to collect the PM. Afterwards, the PSDs were measured by a scanning mobility particle sizer (SMPS). Because of the assumption of spherical particles in the simulations, the raw experimental data were corrected regarding their agglomerate sizes by applying the

Conclusion

In this study, the Population Balance Model (PBM) was applied for Particle Size Distribution (PSD) calculations in combination with two different pyrolysis models. To calculate the precursors of the PM, it is assumed that the precursor release is proportional to the volatile release. The volatile release calculated with the CPD model starts later than predicted in the FZM model. This leads in the PSDs, to a shift to smaller PM diameters caused by the reduced residence time of precursors (Na and

CRediT authorship contribution statement

Christian Axt: Conceptualization, Investigation, Writing - original draft. Stefan Pielsticker: Conceptualization, Methodology, Investigation, Writing - original draft. Thobias Kreitzberg: Writing - review & editing. Oliver Hatzfeld: Project administration. Qi Gao: Methodology. Shuiqing Li: Supervision. Reinhold Kneer: Writing - review & editing, Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to thank the German Research Foundation (DFG) (Grant No. 392429716) and the National Natural Science Foundation of China (NSFC) (Grant No. 51761135126) for funding the joint Sino-German research project “Inorganic fine particulate matter formation during turbulent pulverized coal combustion”.

References (47)

  • L. Yan et al.

    The implication of mineral coalescence behaviour on ash formation and ash deposition during pulverised coal combustion

    Fuel

    (2001)
  • L.B. Clarke

    The fate of trace elements during coal combustion and gasification: an overview

    Fuel

    (1993)
  • W.P. Linak et al.

    On trimodal particle size distributions in fly ash from pulverized-coal combustion

    Proc Combust Inst

    (2002)
  • M. Neville et al.

    Vaporization and condensation of mineral matter during pulverized coal combustion

    Symp (Int) Combust

    (1981)
  • C.M. Lee et al.

    Modeling the vaporization of ash constituents in a coal-fired boiler

    Proc Combust Inst

    (2000)
  • M. McNallan et al.

    The formation of inorganic particulates by homogeneous nucleation in gases produced by the combustion of coal

    Combust Flame

    (1981)
  • P.J. van Eyk et al.

    The release of water-bound and organic sodium from loy yang coal during the combustion of single particles in a flat flame

    Combust Flame

    (2011)
  • S.E. Pratsinis et al.

    Particle formation in gases: a review

    Powder Technol

    (1996)
  • H. Zhao et al.

    A population balance-monte carlo method for particle coagulation in spatially inhomogeneous systems

    Comput Fluids

    (2013)
  • Q. Gao et al.

    Measurement and numerical simulation of ultrafine particle size distribution in the early stage of high-sodium lignite combustion

    Proc Combust Inst

    (2017)
  • L. Yan et al.

    A mathematical model of ash formation during pulverized coal combustion

    Fuel

    (2002)
  • W. Fu et al.

    A general model of pulverized coal devolatilization

    Fuel

    (1989)
  • Y. Yuan et al.

    The transition of heterogeneous–homogeneous ignitions of dispersed coal particle streams

    Combust Flame

    (2014)
  • Cited by (0)

    View full text