A Novel Approach to Predicting NOx Emissions From Dry Low Emissions Gas Turbines

Abstract

An empirical modeling concept for the prediction of NOx emissions from dry low emissions (DLE) gas turbines is presented. The approach is more suited to low emissions operation than are traditional approaches (, 1998, Taylor and Francis, New York). The latter, though addressing key operating parameters, such as temperature and pressure drop, do not address issues such as variation in fuel/air distribution through the use of multifuel stream systems, which are commonly applied in DLE combustors to enable flame stability over the full operating range. Additionally, the pressure drop dependence of NOx in such systems is complex, and the exponent of a simple pressure drop term can vary substantially. The present approach derives the NOx model from the equations that govern the NOx chemistry, the fuel/air distribution and the dependence of the main reaction zone on its controlling parameters. The approach is evaluated through comparing its characteristics to data obtained from high-pressure testing of a DLE combustor fueled with natural gas. The data were acquired at a constant pressure and preheat temperature (14 Bara and 400°C) and a range of flame temperatures and flow rates. Though the model is configured to address both relatively fast and slow NOx formation routes, the present validation is conducted under conditions where the latter is negligible. The model is seen to reproduce key features apparent in the data, in particular, the variable pressure drop dependence without any ad hoc manipulation of a pressure drop exponent.