With the global warming, many countries pay more attention to environmental pollution. The NOx emissions has become an important index when gas turbine designed. This paper provides a method for predicting NOx emissions of marine gas turbine under variable operating conditions. Firstly build the 3-D model of combustor. The characteristic regions of combustor were divided according to the reaction principle. Then build the chemical reactor network (CRN) models of different characteristic regions. According to the NOx emissions of several specific operating points simulated by computational fluid dynamics (CFD), fit the relation between residence time and operating conditions by Newton interpolation in the CRN models. Then the prediction model of NOx emissions of gas turbine was established by using neural network. The NOx emissions under 0.7∼1.0 working conditions and 0.019∼0.023 fuel-air ratios can be predicted efficiently.

This content is only available via PDF.
You do not currently have access to this content.