Improved process understanding and optimization by multivariate statistical analysis of Microbial Fuel Cells operation

The aim of this work is to analyze Microbial Fuel Cell (MFC) processing of dairy wastewater with a multivariate statistical approach. An operating MFC was monitored for 70 days using dairy influents with varying characteristics. Results of a Principal Component Analysis (PCA) suggested that the initial dataset of 8 process-related variables could be reduced to 3 main components, explaining 80% of the cumulative variance. The first principal component (PC1) was strictly related to the conductivity of the influents and the performance of the MFC (in terms of COD removal and CE), while PC2's main contributors were: influent pH, power density and COD of the anolyte. Finally, PC3 was related to the anolyte characteristics (pH, CODin) and CE. Results describe how relationships between operational variables can lead to the definition of new sets of explanatory variables to improve process visualization and to further process modifications for its optimization.