Determination of the lower calorific value on the lignite coal by using artificial neural networks

The lower calorific value of the coal is analysed by using the coal's moisture, volatile matter, ash and sulphur. From this analysis, heat and sulphur values are determined by instrumental analysis techniques. However such analysis process negatively affects businesses both in terms of cost and time. In order to prevent such disadvantages, particular heating value can be determined by a more simple method by using ash and moisture values. In this study, the laboratory analyses of tuvenane coals produced from open pit and undergraund areas of Garp Lignite Enterprise Management operating under General Directorate of Turkish Coal and tuvenane coals drilled from mine areas of Çelikler Seyitömer Lignite Enterprise Management, are investigated by back propagation algorithm of artificial neural network. With the idea of coal's moistrure, volatile matter, ash and sulfur values can be used determination of the the lower calorific value and in this wise time and cost can be reduced, an ANN(Artificial Neural Network) application is developped in C# software language. With the application, the determined lower calorific values and the real the lower calorific values got from laboratory analyses are compared. When lower calorific values found by the developped ANN model and lower calorific values found by laboratory analyses are compared, it is achieved that 0,972 level of correlation and a significant relationship between values as statistical(p<0.001). According to the results, as an alternative to laboratory analyses, developed software can be used to determine the lower calorific analysis of coal.