Session: 04-31 Emissions I

Paper Number: 127627

127627 - Predicting Gas Turbine NOx Emissions With Machine Learning 

Gas Turbine operations result in production of NOx and CO, both of which are pollutants that are to be strictly monitored and controlled. To control NOx within permissible limits, Combustion Emissions Monitoring System (CEMS) readings are to be continuously monitored and flagged for any anomaly in NOx readings. This paper covers the method for developing a proof of concept of a Machine Learning (ML) model that predicts NOx from a fleet of General Electric 7FA (GE7FA) and Siemens Westinghouse 501F (SW501F) sites monitored. The foundation of this project used the open-source Keras library and machine learning framework TensorTlow to construct, develop, test, and tune all respective models in Python. The authors will show that working with quantitative, time-series data, a deep neural network (DNN) using a rectified linear unit activation function was deemed the most appropriate for the given issue at hand. Using this framework in conjunction with the domain knowledge from a Monitoring & Diagnostic team and gas turbine experts, a DNN for each frame type was successfully constructed. The 7FA test site and SW501F test site produced NOx predictions with a Mean Absolute Error of 0.45 and 0.7 ppm respectively, hence demonstrating the capability of a machine learning model to accurately predict NOx emission for any typically trained scenario.

Presenting Author: Harrison Eller Power Systems Mfg.

Presenting Author Biography: An aspiring engineer & data analyst looking to revolutionize the industry using big data. Recently graduated from the University of Tennessee with a bachelor's degree in mechanical engineering and a master's degree in business analytics. Specializes in handling large sets of technical data to proactively monitor combustion turbines across the globe.

Authors:

Harrison Eller Power Systems Mfg.
Sanjana Singh Power Systems Mfg.
Sumit Soni Power Systems Mfg.