NOx Prediction and Control in Oxygen Enriched Air Natural Gas Burners

Air Products has an ongoing program to develop oxygen-based technologies for combustion processes used in heating and melting applications. These processes are typically found in the glass and metals industries. Oxygen enrichment of existing air-fuel combustion processes can increase production, reduce fuel consumption, and reduce flue gas volume. As part of this program, a model is being developed to preqict the impact of oxygen enrichment on NOx levels in the flue of industrial furnaces. The model (GRREK) incorporates gas flow and radiation with reaction equilibrium and kinetics. GRREK is being used to predict flue NOx at various oxygen enrichment levels in a furnace. It is also being used to identify the important parameters in the production and control of NOx in order to develop alternatives to the capitally intensive post combustion cleaning systems. GRREK is designed for steady-state, turbulent flow in a two-dimensional cylindrical furnace with a burner at one end and an exit flue at the other end. The full Navier-Stokes equations are solved using a finite difference, control volume technique. The model is designed to solve any level of complexity for the NOx chemical kinetics. Each control volume in the furnace is treated as a Perfectly Stirred Reactor (PSR). Radiative transport between the flame and the furnace is modeled using the method of discrete ordinates. The paper compares the model predictions of NO against data taken in the Air Products combustion furnace for a North American air-fuel burner tested at various stoichiometries and levels of oxygen enrichment. The measured values at two stoichiometries are compared against the theoretically predicted values. The model predicted the correct trend, but an order of magnitude higher than the measured values. The laboratory tests and model development are being jointly funded by the Gas Research Institute and Air Products.