Comparative analysis and prediction study for effluent gas emissions in a coal-fired thermal power plant using artificial intelligence and statistical tools

Abstract

Environmental sustainability is a crucial phenomenon for daily human life & planet earth, where various plants influence a massive impact to atmosphere as a drawback of the industrial revolutions. The worldwide regulations imposes various improvements in existing coal-fired thermal power plants by means of implementing emission reduction systems, desulphurization, de-NOx systems, modernization of instrumentation, sensors and automation control systems, co-firing process, efficiency enhancements, etc. which cause high investment impact to the plants. Therefore if relevant variables those affecting emission rates are selected, and the emission measures are predicted and controlled properly, then it will be beneficial to implement emission reduction scheme to the plant accordingly. In this research, effluent gas emissions of a 180 MW coal-fired thermal power plant located in Kocaeli, Turkey is modelled and predicted using Artificial Neural Networks (ANN), Autoregressive Integrated Moving Average (ARIMA) and Multiple Linear Regression (MLR) approaches within proposed process parameters. The data is collected from the consolidated continuous emission monitoring system reports during 1 month of operation, and proposed performance criterion, regression coefficient and root mean squared error, is used for comparison of the model outputs. Prediction results are found satisfactory and encouraging in order to integrate the models to the power plant automation systems as an expert system.