A Municipal Solid Waste Heating Value Predicting Model Based on Artificial Neural Network

Proper management of waste produced by human activity has been one of the major environmental concerns of modern societies. The use of the waste as a source of renewable energy is an interesting option, increasing the country's energy matrix, reducing the volume of waste and mitigating its toxicity. However, for a proper design and operation of a waste-to-energy plant, it is essential to know the energy potential of the waste. This paper aims to develop a computer model to predict the higher heating value (HHV) of the municipal solid waste (MSW), based on the ultimate analysis, using an artificial intelligence technique known as artificial neural networks (ANN). With the use of ultimate analysis from the literature, we tested different ANN architectures until the best architecture was found. We performed comparative tests with mathematical models from the literature, and the ANN model got a considerably more accurate response than the mathematical models analyzed. The ANN model showed a superior precision in terms of mean absolute percentage error (MAPE = 2.9%) and in terms of the coefficient of determination (R2 = 0.996). Finally, we conclude that the forecasting of the HHV of the MSW could be improved through the use of ANN models, and the proposed model is very suitable to the task.