1. Introduction

The optimum design and operation of pulverised coal burners requires a detailed knowledge of the fuel properties. In the last 15 years the IFRF Isothermal Plug Flow Reactor ([GLOSS]IPFR[/GLOSS]) has been used to develop techniques for the characterisation of coal combustion behaviour under industrial conditions. The concept of characterisation introduced in CF48 and the use of the IPFR as an advanced characteristic technique is introduced in CF135.

Such characterisation allows the prediction of coal burnout at any process condition (temperature and oxygen concentration) in the reactor. Coal is burnt in an environment similar to that found in pulverised coal combustion systems but under the well controlled conditions of IPFR as described in CF135.

The combustion of a coal particle can be divided into two separate processes, devolatilisation and the char burnout. The devolatilisation step can occur within 5 to 100ms, where the majority of the volatile components in the coal vaporise and escape from the coal structure. This leaves a lattice of carbon and ash, which can burn for up to 6 seconds before ideally being transformed completely to ash.

In this combustion file we consider the high temperature devolatilisation characteristics.

In pulverised coal combustion, the relatively high heating rate experienced by coal particles and the high maximum temperature reached gives rise to a volatile release generally higher than that found in proximate analysis.

To demonstrate how the IPFR may be used to measure high temperature volatile release, the results of the analysis of a Colombian coal, El Cerejon, are used. This coal has been characterised on three separate occasions [1], and all three sets of results are used.

2. High Temperature Devolatilisation

Although the devolatilisation process occurs in a very short time, in the region of 5 to 100 ms, it plays a large part in determining the burning characteristics of the particle. Devolatilisation is where some of the components in the coal transform to a gaseous state on contact with heat energy, causing the carbon structure to open out into a honeycomb configuration as the newly formed gas escapes. The speed and extent of devolatilisation is dependent on both the combustion conditions and the physical nature of the coal.

The standard volatile matter content of a coal is given by the [GLOSS]Proximate analysis[/GLOSS] – [VM]prox. Typically, the High Temperature Volatile Matter ([GLOSS]HTVM[/GLOSS]) content [VMmax] can be considerably higher than that given by the proximate analysis. From these, we can define a [GLOSS]HTVM Yield factor[/GLOSS] as:

Sayre et. al., [IFRF Doc No F 88/y/13] demonstrated that the HTVM Yield Factor could vary from values form 1.1 to 1.8 and that there was little or no relation ship between the yield factor and the volatile matter content given by the proximate analysis – See Figure 1

Figure 1: Relationship between proximate volatile yield and high temperature Yield Factor

Thus, is clear that in order to predict the flame and combustion characteristics of a pulverised coal, it is important to measure the HTVM release – [VMmax].

Furthermore, the rate of volatile release can affect the char reactivity. Of prime importance for the devolatilisation is the heating rate. This dictates the speed that the volatile components become gaseous, and also the size of pores in the carbon. The carbon shell has tiny pores, vesicles, allowing access to the inside of the sphere. Generally the more pores that are opened in the devolatilisation process, the more the availability of reactive surface area, allowing a better burnout to be achieved.

3. HTVM Release – El Cerejon Coal

The proximate volatile matter [VM]prox of the three samples of El Cerejon varies from 32% to 39% mass (d.a.f.)

Figure 2 shows the weight loss during devolatilisation measured in the three different measurement campaigns. The measurements were all done under similar conditions in an oxygen free environment at 1400C.

The weight loss (HVTM release) is determined by the ash tracer technique of the ash content of coal and char. This is determined through the equation:

For El Cerejon coal, this volatile matter varies form 52% to 58% volume (d.a.f.). Therefore the VM Yield factor varies between 1.35 and 1.45. This confirms a larger error range for the standardised proximate analysis compared to the HTVM yield from the Isothermal Plug Flow Reactor, see also figure 2

Figure 2: weight loss during devolatilisation of El Cerejon Coal

4. Conclusions

The variation on HTVM Yield Factor measured for a wide range of coals demonstrates that it is not possible to predict this value with reasonable accuracy. Therefore, for detailed design of burners and combustion systems, data of the type generated on the IPFR is very important.

Sources

[1] G. Gallagher, J. Haas, Reference report: Characterisation of weight loss and char burnout behaviour for El Cerejon Coal, IJmuiden, The Netherlands, 1996. IFRF Doc No 960927.
[2] A.N. Sayre, K.J. Knill and J.P. Smart, Research report: Coal characterization requirements for modelling pulverized coal flames, IJmuiden, The Netherlands, 1996. IFRF Doc No F 88/y/13.

Acknowledgements

The author would like to thanks G. Gallagher and J. Haas for the information supplied and for the work carried put in the IFRF related with the Isothermal Plug Flow Reactor