1. Background

The difficulties in choosing NOx sub-models for oxy-coal combustion modelling using [GLOSS]RANS[/GLOSS] are largely the same as those faced when modelling air-fired coal combustors. NOx formation can be divided into a small number of paths that in many cases can be considered separately. These are: formation of NOx through the thermal path; the prompt-NO path, occurring through reaction between hydrocarbon radicals and N₂; and the fuel-NO path, describing the fate of nitrogen containing species released with volatiles. In addition, the nitrogen remaining in the [GLOSS]char[/GLOSS]also needs to be taken into account.

Nevertheless, the sub-models applicable to modelling of air-fired conditions are not necessarily applicable to oxy-coal combustion. The reason for this is the enhanced CO₂ level that alters the radical pool, mainly by increasing the level of OH-radicals. Since most sub-models rely on empirical reaction rate descriptions, these may need to be revised. In most cases the sub-model for prompt-NOx formation can be omitted, since the relative contribution from this path is likely to be within the error limits. 

In general, a simplified description of the complex nitrogen reactions should be used with caution since they may not be reliable, as pointed out by Hansen and Glarborg [1].

Sub-models describing the NOx chemistry are typically run as post-processing, which reduces the required computational effort. Nevertheless, there is probably no model that is capable of fully describing the complex interaction between turbulence and chemistry for RANS modelling. This too adds to the uncertainties occurring in NOx modelling. 

2. Thermal-NOx

A limited set of elementary reactions is sufficient to describe the formation of thermal-NO. Those reactions describing the elementary reactions involving nitrogenous species are often referred to as the (extended) [GLOSS]Zeldovich mechanism[/GLOSS]. This mechanism could principally be used directly, but the reaction rates also depend on the concentration of a small number of key radicals. These can be estimated in various ways. Suggestions on how model thermal-NOx in general can be found in CF40. However, due to the low N₂ levels in oxy-coal combustion the reverse reactions may be important [2]. Therefore a thermal-NOx sub-model that also accounts for the destruction of NO is recommended.

Thermal-NOx formation is sensitive to temperature fluctuations. Consequently the sub-model should account for this. Often a [GLOSS]pdf[/GLOSS]-based approach is used for this purpose.

3. Fuel-NOx

For the sub-model describing the conversion of fuel-NOx the modeller has to rely on the same sub-models as used in modelling of air-fired cases. CF41 deals with fuel-N chemistry. The sub-model describing the conversion of fuel-NOx typically consists of a limited number of empirical reactions. There also exists a reduced mechanism for describing fuel-N conversion in coal-combustion [3]. Although in reality the higher level of CO₂ prevailing at oxy-fired conditions further increases the uncertainty due to the afore-mentioned higher OH-radical level and the higher three-body efficiency of CO₂, the dominant uncertainty stems from the modelling of the volatile release from the coal. Problems here include how much of the nitrogen in the coal is released during devolatilisation, at which rate it is released and in which form? The most common assumption is to assume that most of the fuel-N is released as HCN and at a rate that is proportional to the rate of release of other volatiles. An important difference to air-fired conditions is that NO is also recycled. This does not have to be to be taken into account separately as a proper model for fuel-NOx also includes reactions between NOx-precursors and NO.

4. Reburning chemistry

Reburning, i.e., the use of a secondary reducing zone where already-formed NOx is converted to HCN through reactions with radicals of the form CH_i, is a commonly applied technology to abate NOx formation. This can be done on a boiler scale (CF130) or on a burner scale (CF127). There exists a number of simplified schemes for NOx reburning, but here too, the warning given earlier that simplified descriptions should be used with caution applies. The reactions describing the reburning chemistry should be included in the sub-model describing the fuel-NOx. This sub-model should be able to account for the fluctuations in the species field. Models based on systematic reduction of full kinetic schemes have also been developed [4]. Using such a scheme in a sub-model may prove impractical as the algebraic expression involved may be numerically problematic.

5. Char-NOx

In addition to the homogenous reaction, i.e., the pure gas phase chemistry, heterogeneous reactions also occur in coal combustion. In this case, too, the same sub-models used to model NOx chemistry, where reactions with char is involved, can also be used in oxy-coal modelling, although gasification reactions should be considered to be of principal importance. There are two main reaction paths that need to be considered: one is the oxidation of char-N, the second is reduction of gas-phase NO by heterogeneous reactions with the char. For the latter, the empirical expression rate suggested by Garijo et al. [5] has been validated for oxy-coal conditions by Hashemi et al. [6].

6. Conclusions

In most cases, the only feasible option is to use the same sub-models for NOx modelling in oxy-coal cases as in the modelling of air-fired cases. Two issues needs to be taken into account: 1) that the recirculation of NO makes reactions with NOx precursors important and 2) that due to the low N₂ concentration, the backward rates in the Zeldovich mechanism may become important. Finally, a full description of NOx chemistry requires an extensive set of elementary reactions. A simplified description of the chemistry should be used with caution.

Sources

[1] Hansen, Stine and Peter Glarborg. “A Simplified Model for Volatile-N Oxidation”. Energy & Fuels. 2010, 24. 2883-2890.

[2] Normann, Fredrik, Klas Andersson,  Bo Leckner and Filip Johnsson. ” High-temperature reduction of nitrogen oxides in oxy-fuel combustion”. Fuel 2008, 87, 3579-3585.

[3] Pedersen, Lars Saaby, Peter Glarborg, and Kim Dam-Johansen. “A Reduced Reaction Scheme for Volatile Nitrogen Conversion in Coal Combustion”. Combustion Science and Technology. 1998, 131(1-6). 193-223.

[4] Glarborg, Peter and Stine Hansen. “Simplified Model for Reburning Chemistry”. Energy & Fuels. 2010, 24(8). 4185-4192.

[5] Garcia Garijo, M`.Elena, Anker Jensen, and Peter Glarborg. “Reactivity of coal char in reducing NO”. Combustion and Flame. 2004, 136(1-2). 249-253.

[6] Hashemi, Hamid, Stine Hansen, Maja Toftegaard, Kim Pedersen, Anker Jensen,Kim Dam-Johansen and Peter Glarborg. ”A Model for Nitrogen Chemistry in Oxy-Fuel Combustion of Pulverized Coal”. Energy Fuels 2011, 25, 4280–4289.

Acknowledgements

RELCOM; Reliable and Efficient Combustion of Oxygen/Coal/Recycled Flue Gas Mixtures.

Project undertaken with the financial support of the European Commission

FP7 Grant Agreement Number: 268191.