Sustainable Water Management for Steam Generation in Sugarcane Biorefineries: Applying PCA and MST Clustering in Sample Analysis

Abstract

Our study on sustainable water management in sugarcane biorefineries, which utilizes water as a primary resource for generating bioenergy through steam production, has employed a novel approach. High water quality is crucial for optimal efficiency, particularly in boiler operations. We have utilized unsupervised methods, such as principal component analysis (PCA) and minimum spanning tree (MST), alongside instrumental analysis data, to assess water quality in steam production. The PCA exploratory analysis identified three distinct clusters, with the relevant variables being conductivity and SiO2 content, to differentiate the purity of a dataset of 120 samples. MST-based clustering corroborated the PCA findings, forming three clusters: sample 1 represented the purest water, while samples 3 and 6 were in different clusters, indicating less purity in boiler feedwater. These unsupervised methods are highly effective, providing accurate and reliable data analysis and significantly benefiting sugarcane biorefineries by eliminating subjective biases. The findings of this study promise to improve water management practices in sugarcane biorefineries, leading to more efficient and sustainable operations.