Improved Techniques for the Prediction of NOx Formation from Char Nitrogen

This study has shown that under staged-air conditions (such as those in low-nox burners) commonly now used to reduce emissions, the ultimate NOx emissions will be, to a large extent, determined by the char-n.  The results show that around half of the char nitrogen will be converted to NOx. Hence, for coals of equivalent coal-N content, those which release more of their nitrogen with the volatiles have the potential to produce lower NOx emissions than those from a coal of low volatile content. This study has been successful in demonstrating that coals with high N content (including many Australian coals) do not necessarily produce higher NOx emissions under conditions used in modern power plants burning coal.  All combustion processes which utilise fossil fuels result in the formation of oxides of nitrogen (NOx), which are significant contributors to a number of detrimental atmospheric environmental effects.

In modern power plants burning coal, greater than 80% of the NOx can be formed from the nitrogen contained in the fuel. The nitrogen content of Australian black coals generally falls in the range from 1.5-2%(daf) with a number of coals having greater than 2% nitrogen. These coals have experienced marketing difficulties despite evidence that NOx formation from these coals is not excessive.

Project Objectives

Project Aim

The aim of this project was to reduce the negative market perception of the high N content of some Australian coals through an improved understanding of the reactions of char-N. This was achieved by determining the proportion of char-N converted to NOx over a wide range of conditions.

Preparation of char

A variety of NSW and Queensland coals were used for both fixed bed reactor and entrained flow reactor experiments. The nitrogen content of the coals ranged from 1.1 - 2.2 wt% (daf basis).

Chars produced under oxidative pyrolysis conditions and created under high heating rate conditions, were more likely to approximate chars formed at full scale in a coal fired furnace, than chars produced in a fixed bed reactor.  These chars were characterised by low volatile matter (VM), low hydrogen and oxygen content and N/C ratios similar to those in the parent coals.

Fixed Bed Reactor Experiments

The captive study of char-N conversion to NO was conducted at the University of Sydney in a fixed bed (TGA) reactor at temperatures between 600 and 1100°C with different O2 concentrations (1.1, 2,4 and 5.4%). Partial oxidation experiments, designed to investigate changes in the nitrogen composition during oxidation, were also conducted.

Included in this study was computer kinetic modelling of the gas phase reactions which were likely to occur within the TGA system as well as the modelling of the heterogeneous kinetics. The program used calculated species and temperature profiles for the reaction of a porous solid in a reacting gas environment at pseudo-steady state.

It was used to model the oxidation of coal char in the TGA with the intention of gaining insights into the nature of the chemical processes that determine the fate of the char nitrogen.

Entrained Flow Reactor Experiments

Char combustion experiments were conducted at CSIRO in an entrained flow reactor with a gas temperature of 1200±50°C. These experiments were designed to determine the effect of the combustion stoichiometry on the conversion of char-N to NO. Equivalence ratios of between 0.6 and 1.75 were used for this study.

These experiments were unique in that char-N release was determined by direct measurement of N2 and NO which was made possible by using an O2/Ar reaction environment.

Project Outcomes

The review of the existing techniques for predicting NOx emissions from the combustion of coal revealed the following points:

• formation of NOx in coal combustion is complex, and depends on boiler design and   operating parameters as well as on coal characteristics; there appears to be no simple, direct relationship between coal-N content and NOx produced;

• combustion modifications, such as air- and fuel-staging, result in lower NOx emissions for all coals to a varying degree;
•  recent studies show that the effects of coal quality on NO emissions under the low NOx regime produced by low NOx burners relate to the nitrogen which is retained in the char, implying that coals which release most of their nitrogen with the volatile matter may produce lower NOx emissions;

In the fixed-bed (TGA) reactor experiments, where the evolution of char-N was studied as a function of oxidation temperature, oxygen concentration and fractional char conversion revealed that:

• char combustion results in the formation of N2 and NO as the major nitrogenous  products with small concentrations of N2O and HCN detected at temperatures below 830°C. The HCN is a primary product of char oxidation;

• the fraction of char nitrogen exiting the reactor as NO is strongly dependent on the oxidation temperature;
• char combusted at low temperatures is initially enriched in nitrogen, with this nitrogen being released during the final stages of char burnout. XPS shows that this enrichment is a surface effect.

Char combustion experiments performed in an entrained flow reactor, under conditions approaching those found in p.f. combustors, have shown that:

• for equivalence ratios in the range from 0.6-1.75, char-C and char-N were released at similar relative rates over the range of carbon burnout from 20-80%;

• for equivalence ratios of around 1, the fractional conversion of char-N to NOx is 0.5±0.1; under fuel-rich conditions lower fractional conversion was observed, and under fuel-lean conditions higher fractional conversions were observed;
• there was little influence of the parent coal or its nitrogen content.

A modelling study of char combustion incorporating gas-phase and heterogeneous reactions and mass-transfer effects was conducted. This model showed that the major features of both the TGA and the entrained flow experiments were consistent within a single modelling framework and that the apparent differences between the NO yields in the two experiments are explicable in terms of the different experimental arrangements.