An Experimentally Verified NOx Prediction Algorithm Incorporating the Effects of Steam Injection

Abstract

A Perfectly Stirred Reactor model, kinetic rate data, and equilibrium adiabatic flame temperatures have been incorporated into a closed form NOx prediction algorithm. The algorithm accounts for combustor inlet parameters and steam injection. Combining this model with the operating map for the gas turbine allows prediction of NOx with variation in parameters such as inlet guide vane angle, ambient temperature, and load. If the actual values of the model input variables are measured, real time prediction of NOx emissions may be generated using a microcomputer. This signal may then be used as an input to NOx abatement systems such as Selective Catalytic Reduction. The semitheoretical technique is based upon the extended Zeldovich chain reaction kinetics for the production of NOx in the post-flame zone. The temperature and concentration of major product species in the post-flame zone are taken to be those appropriate to equilibrium. Steam injection and inlet effects enter the model through their influence on the abiabatic equilibrium flame temperature. Mixing effects are accounted for empirically.