NO X ABATEMENT FROM OIL COMBUSTION – REBURNING AND LOW NOX BURNER Renato Vergnhanini Filho IPT - Technological Research Institute of São Paulo State, São Paulo, SP, Brazil vergnhan@ipt.br Francisco D. A. de Sousa *; Carla S. T. Marques and Celso A. Bertran **; Ricardo Serfaty *** * IPT - Technological Research Institute of São Paulo State, São Paulo, SP, Brazil ** Unicamp - State University of Campinas, Campinas, SP, Brazil *** Petrobras - Brazilian Petroleum Company, Brazil Abstract. The performance of a low NO x burner and a reburning system for heavy oil combustion has been investigated experimentally. Reburning and burnout simulations were carried out initially in order to determine preliminary experimental parameters for maximum NO x reduction. The kinetic model was later used to simulate the experimental data in the conditions that showed higher NO x reduction. The main conclusions are: (1) the implementation of reburning yields a reduction of 63-71% (according to stoichiometric ratio in the reburning zone) in the NO x emission related to the unstaged fuel value, without detriment of combustion quality; (2) the implementation of air staging by low NO x burner yields a reduction of 48% in the NO x emission related to the unstaged value, however the particulate material (PM) emission rates tended upwards; (3) the implementation of reburning in a low NO x flame yields a reduction of 80% in the NO x emission related to a base line flame (unstaged fuel and air), however the PM emission rates tended upwards and (4) an unique kinetic model is unable to make a good simulation of reburning and burnout zones even if the injection systems provides good mixing between the primary gas stream, the natural gas and the supplementary air, showing the necessity of a mixing-kinetic model. Keywords. Low NO x burner, reburning system, heavy oil combustion, reburning and burnout simulations, kinetic model 1. Introduction Fuel oil is an important industrial fuel in Brazil, as it participated as 10.6% of the energy consumption in industries in agreement with the 2001 national energy balance. Brazilian fuel oils are produced by Petrobras (Brazilian Petroleum Company) from RASF (asphaltic residue) and RESVAC (vacuum residue) diluted with light fractions at different proportions and having different specifications which differ basically in viscosity and sulfur content. These fuel oils are classified as types 1 to 9 according to their increasing viscosity and as A or B according to their sulfur content, A for the high (above 1%) and B for low one (up to 1%). They have abnormally high viscosity (the type 1 is more viscous than United States residual oil, referred to as Bunker C) and nitrogen content (usually around 1%) and in the burning they tend to produce flue gases with high content of particulate material (PM) and nitrogen oxides (NO x ). The performance of a low NO x burner and a reburning system for heavy oil combustion, both designed and built by IPT, was investigated in a horizontal test furnace. Natural gas was used as reburning fuel. Furthermore, a kinetic model was built in order to estimate the initial experimental parameters to higher NO reduction and to obtain a better description of reburning and burnout zones for the trials carried out. This paper reports the activities achieved and the results obtained. 2. Experimental set-up 2.1. Test furnace The experimental work was carried out in a cylindrical horizontal test furnace of 1.1 m internal diameter and 4.0 m length. It has an internal refractory lining that allows internal surface temperatures to be maintained around 900 o C. The furnace steel shell is double plated and divided in nine sections. Each section, as well as the furnace front end are provided with independent water cooling flows. This allows the computation of the net heat flux distribution on the furnace internal surface and the fraction of the total power transferred to it. The combustion chamber is provided with a longitudinal opening where measurement probes can be inserted. The furnace is coupled to a horizontal channel. This channel is attached to a air preheater, which is coupled to a stack provided with a platform where the sampling station for gas composition and PM collecting is located. In all trials the furnace internal pressure was kept at 25 Pa (effective) to avoid air incoming through leakages. 2.2. Burner A burner was designed, according to the scaling criteria (constant residence time approach combined with geometric and thermal similarities), and built by IPT to be coupled to the test furnace. The burner can operate either as a