Development and selection of lignocellulose biomass and nano-additive combination for co-pyrolysis operation in power generation using hybrid prediction and Machine learning model – A k-means cluster approach

Abstract

The burning of carbon rich fuels is associated to be the primary cause of developing large quantity of greenhouse gases which alters the earth’s ecosystem, thereby causing problems in human health and environment. This study investigates the use of various lignocellulose biomass sources—Sugarcane Bagasse, Rice Husk, Wheat Straw, Moringa, and Vetiver—in conjunction with different nanoparticles to create combinations aimed at reducing volatile matter to be used in co-pyrolysis operation in a thermal powerplant. The chosen outcome parameters for clustering these nanoparticles include energy yield, combustion efficiency, ash generation, and SO₂ emission. An Adaptive Neuro Fuzzy Interface System (ANFIS) model is employed to identify trends and relationships between biomass-nanoparticle combinations and the output parameters. K-means cluster analysis is used to categorize the combinations into best, worst, and average clusters. The ANFIS algorithm reveals that the relationship is Trapezoidal with the smallest combined error rates. Among the tested combinations, Moringa coupled with silver nanoparticles emerged as the optimal biomass-nanoparticle pair, exhibiting the smallest centroidal distance of 0.14 from the best cluster centroid. Moringa and Silver nanoparticles achieved significant cost reduction and emissions reduction, with outputs showing an energy yield of 19.8 Mg/kg, combustion efficiency of 82.4 %, ash generation of 2.7 %, and SO₂ emission of 0.2 g/kg.