Data-driven correlations of higher heating value for biomass, waste and their combination based on their elemental compositions

The biomass and waste solid fuel properties, higher heating value (HHV), are the important factor contributing to the optimization studies towards efficient thermochemical conversion processes. In this study, a new data-driven correlation models were developed for assessing the HHV of biomass and waste solid fuels as a function of the ultimate analysis or elemental compositions. An artificial neural network (ANN) method was used to connect the inputs and the outputs information thorough iterative training procedure. The models were developed and tested using 930 biomass datasets and 500 waste datasets obtained from the published literature. The commercial programming software was employed with the proper, 5-25-1 ANN architecture. The prediction ability of the obtained models was compared with those testing data with high accuracy. Accordingly, the obtained data-driven correlations were capable of predicting excellent results for all the biomass, waste and their combination datasets. In addition, the developed data-driven correlation model for the biomass-waste combination dataset was accurately enough for generally representing the solid fuel prediction. Finally, the values of weights and biases for the developed data-driven correlation model of biomass-waste combination dataset were summarized. The findings may help to provide an appropriate design and operation of the thermochemical conversion process. (c) 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of the 2nd International Conference on Power Engineering, ICPE, 2021.